Abstract
Summary
Attenders to fracture liaison services in a quality assessment study of 1936 women and men ≥ 50 years had 20% lower risk of subsequent fractures and 33% lower mortality than non-attenders. Patient’s who were adherent to AOD after 12 months had a 49% lower risk of fractures than non-adherent patients.
Purpose
Secondary fracture prevention rates are low in Norway and internationally. To overcome the existing treatment gap, fracture liaison services (FLS) are recommended. The study aimed to assess impact of FLS on fracture rates and mortality among attenders versus non-attenders, and patients adherent versus non-adherent to AOD.
Methods
This single-center sub-study of the Norwegian Capture the Fracture Initiative included 1936 women and men ≥ 50 years of age with an index fracture of any type during 2015–2018. We estimated hazard ratios (HR) for subsequent fractures and mortality in 1041 attenders versus 895 non-attenders to the FLS, and 427 patients who were adherent versus 174 patients who were non-adherent to anti-osteoporotic drugs (AOD) 12 months after treatment initiation.
Results
During a median 3.5-years observation (range 0.5 to 6.3), attenders had a 20% lower risk of subsequent fractures of any type (HR 0.80; 95% CI 0.64–0.999) and 33% lower mortality (HR 0.67; 95% CI 0.51–0.89) than non-attenders. Patients who were adherent to AOD after 12 months had a 49% lower risk of subsequent fractures than non-adherent patients (HR 0.51; 95% CI 0.35–0.72), but no difference in mortality was shown. Models were adjusted for age, sex, body mass index, smoking, prior fracture, and index hip fracture.
Conclusion
These results confirm that patients who were able and willing to attend FLS had lower risk of subsequent fractures and mortality than non-attenders. Patients who were adherent to AOD had lower risk of subsequent fractures than non-adherent patients, emphasizing the importance of adherence in secondary fracture prevention.
Keywords: Fracture liaison services, Men, Mortality, Secondary fracture prevention, Subsequent fracture, Women
Introduction
The estimated lifetime risk of sustaining an osteoporotic fracture at the age of 50 is 55% in women, and 25% in men [1]. A hip fracture is the most severe type of osteoporotic fracture with a risk of permanently loss of function and independence, and 25% mortality after one year [2]. It is also one of the most resource demanding diagnoses for healthcare systems and societies [3, 4]. A prior fracture doubles a patient’s future fracture risk, and multiple fractures increase the risk up to five-fold [5]. A range of anti-osteoporotic drugs (AOD) are available to reduce the risk of fractures [6–10].
Despite the high social and personal costs of osteoporotic fractures, secondary fracture prevention is suboptimal [11]. To overcome this care gap, the establishment of fracture liaison services (FLS) is recommended [12]. FLS is led by a coordinator, who identifies fracture patients and allocates them to fracture risk assessment and AOD treatment when indicated [13]. Most studies on the association of FLS with reduced fracture risk and mortality have compared fracture rates before and after FLS was introduced or compared hospitals with and without FLS [12–19].
In the Norwegian Capture the Fracture Initiative Trial (NoFRACT), FLS was associated with 10–13% lower risk of fragility fractures, 21–25% lower risk of hip fractures and 15–18% lower mortality in both sexes [20, 21]. These results were based on national register data including 100,198 patients with an index fracture in a multicenter, stepped-wedge cluster-randomized trial. Individual level data on whether patients had received the FLS intervention or treatment with AOD were not included in the analysis. Information about important confounding factors, such as the body mass index (BMI), was not available.
This study collected novel individual level information from both attenders and non-attenders of the NoFRACT study at the University Hospital of North Norway (UNN), as well as data on adherence to AOD treatment. The aims of this study were to evaluate the effect of FLS on subsequent fracture rates and mortality in patients who were attenders versus non-attenders to the FLS intervention, patients who were adherent versus non-adherent to the AOD prescribed, and the reasons for non-adherence.
Material and methods
Study population
The NoFRACT trial introduced FLS at departments of orthopedic surgery at seven Norwegian hospitals [20, 21]. The main aim was to assess the effect of FLS on subsequent osteoporotic fracture rates and mortality after an index low-energy fracture. The seven sites were randomized for the order of the starting date by the Norwegian Osteoporosis Association and divided into three clusters with a 4-month interval between the clusters [21]. The current site, UNN, Tromsø, introduced FLS in Oct 2015 [21], with an intervention period from Oct 1, 2015, to Dec 31, 2018, and follow-up to Dec 31, 2019. In brief, women and men ≥ 50 years of age with a recent low-energy fracture of any type (except fingers, toes, face and skull) were contacted by a coordinating nurse within 6–8 weeks after the fracture. Very frail patients with short life expectancy in the judgment of the nurse were excluded. Patients were invited face-to-face or by letter. Patients who were able and willing to attend the FLS intervention were defined as attenders, while those who declined or did not respond to the invitation were defined as non-attenders. Attenders were further divided into adherent and non-adherent. Adherent patients were those who reported that they were able and willing to continue AOD treatment at the follow-up phone-call 12 months after the initiation of treatment. Those who reported that they had stopped taking AOD or were unable or unwilling to continue AOD treatment (including zoledronate) because they had become too frail or did not want to continue treatment for other reasons, were defined as non-adherent. The phone-calls were answered by patients, their spouse, home nurse, or nurse at the nursing home. We had no other information source for adherence (e.g. pharmacy registers or primary care).
FLS intervention
The FLS intervention included identification of fracture cases, assessment and treatment of osteoporosis in patients with a recent low-energy fracture [21, 22]. To capture all fracture cases, the nurse searched daily for ICD-10 fracture diagnosis codes in the hospital system and read the patients’ medical records for quality assessment of the location and date of the fracture. Eligible patients were invited to participate in the study during hospitalization or via mail for those who were treated as out-patients. The nurse informed each attending patient about bone fragility and performed an assessment based on blood samples to exclude secondary causes of osteoporosis, bone mineral density (BMD) measured by dual-energy x-ray absorptiometry (DXA) of both hips and spine, and/or calculation of the Fracture Risk Assessment (FRAX) score, which was reported as the 10-year probability of a major osteoporotic fracture (MOF) [23].
After the assessment, the patients received a letter from a physician with a summary of their fracture risk and a prescription of an AOD if indicated. All the patients were given lifestyle advice on physical activity, smoking cessation, fall prevention, and a healthy diet securing an adequate intake of calcium and vitamin D.
Patients with a hip fracture, vertebral fracture or ≥ 2 other low-energy fractures were recommended AOD regardless of FRAX score or BMD T-score. Zoledronate was recommended to patients with a hip fracture, but this was not implemented at the UNN during the first two years of NoFRACT. Patients with other low-energy fractures, and FRAX score of MOF ≥ 20% or BMD T-score ≤ −1.5 at any site were recommended alendronate, zoledronate, or denosumab. Patients who suffered a fracture during treatment or had a BMD T-score ≤ −3.5 were referred to an endocrinologist for consideration of osteoanabolic treatment. A nurse called all patients who were prescribed AOD at 3 and 12 months after initiation of treatment and collected information on AOD adherence [20]. If needed, a physician modified the treatment.
Data collection
Outcomes were the first subsequent fracture of any type and death. The type of fracture was defined using ICD-10-codes (e.g. S72.0-S72.2 for hip fractures). One of the authors (JK) collected information on the anatomical location and date of prior, index, and subsequent fractures by scrutinizing the patients’ medical record and radiographic reports from the archives at UNN, Tromsø between June 6, 2021, and July 20, 2021. JK also collected data on age, sex, BMI, BMD, FRAX-score, smoking, alcohol intake, glucocorticoid use, rheumatoid arthritis, diabetes type 1 and 2, inflammatory bowel disease, reduced eGFR (< 35 ml/min/1,73m2), stroke, cancer, previous use of AOD, whether patients had an indication for AOD treatment, type of prescribed AOD, and the reasons for not taking AOD at follow-up. Information on co-morbidities was registered from medical records in the hospital system and was partly based on notes made by the NoFRACT nurses and physicians. The data protection officer at UNN recommended this quality assessment study and waived patient informed consent.
Statistical analysis
We performed the student t-test and chi-square test, for comparison of baseline characteristics of attenders vs. non-attenders to the FLS intervention, and patients who were adherent vs. non-adherent to the prescribed AOD 12 months after initiation of treatment. The results are reported as a mean and standard deviation for continuous variables and a number and percentage for categorical variables. To be able to observe any effects it was necessary to have some observation time, and therefore we excluded patients with observation time less than six months. Observation time for fracture outcomes was calculated from six months after the date of the index fracture (during 2015–2018) to the date of first subsequent fracture, death, or end of observation, June 6, 2021. Observation time for mortality outcome was calculated from six months after the date of the index fracture to the date of death, or end of observation, June 6, 2021. Subsequent fracture rates and mortality per 1000 person-years were calculated by dividing the total number of subsequent fractures or deaths by the sum of person-years during observation. Hazard ratio (HR) was calculated using cox proportional hazard models. The HR for subsequent fractures and deaths was calculated in attenders, with non-attenders as the reference, and in patients who were adherent, with non-adherent patients as the reference. The HRs were adjusted for age, sex, BMI, smoking, prior fracture (fracture prior to the index fracture was dichotomized), and index hip fracture. P < 0.05 were considered significant. SAS Software package, v9.4 (SAS Institute Inc., Cary, NC, USA) was used for statistical analyses.
Results
We identified 2024 women and men ≥ 50 years of age with a fracture (Fig. 1). Individuals with pathological fractures (n = 8) or short life expectancy (n = 80) were excluded. This left 1936 patients included in the analysis. Their mean age was 70.0 years (range 50–101) and 73.8% were women. Of 1936 patients, 1041 (54%) were attenders, and 895 were non-attenders to FLS. Of 1041 attenders, 739 (71%) had an indication for treatment with AOD. After 12 months, 601 of 739 were available for follow-up and 138 were lost to follow-up (28 of them died within 12 months). Of the 601patients available at follow-up, 427 (71.0%) were adherent and 174 were non-adherent to AOD.
Fig. 1.
Women and men with osteoporotic fractures included in Norwegian Capture the Fracture (NoFRACT) at the University Hospital of North Norway (UNN) in Tromsø at baseline in 2015–2018. AOD = anti-osteoporotic drugs
Among the attenders, a larger percentage were women (76.6% vs. 70.4%), they were one year younger (69.5 vs 70.6 years), a smaller percentage were smokers (16.5% vs. 24.3%), and a larger percentage had an index hip fracture (23.3% vs. 15.9%) than non-attenders, all p < 0.05 (Table 1). The percentage of chronic diseases did not differ significantly between the groups.
Table 1.
Baseline characteristics of attenders and non-attenders of the Norwegian Capture the Fracture Initiative, at the University Hospital of North Norway, Tromsø
| na | Attenders | na | Non-attenders | p | |
|---|---|---|---|---|---|
| Females, n (%) | 1041 | 798 (76.7) | 895 | 631 (70.5) | 0.002 |
| Age (years) | 1041 | 69.5 ± 10.7 | 895 | 70.6 ± 13.0 | 0.039 |
| Body mass index (kg/cm2) | 858 | 26.1 ± 4.8 | 673 | 26.0 ± 5.2 | 0.741 |
| FRAX score, MOF (%) | 880 | 20.9 ± 11.6 | 633 | 20.7 ± 12.0 | 0.753 |
| Currently smoking, n (%) | 988 | 163 (16.5) | 769 | 187 (24.3) | < 0.001 |
| High alcohol intake, n (%) | 607 | 45 (7.4) | 570 | 56 (9.8) | 0.142 |
| History of previous fracture, n (%) | 1041 | 484 (46.5) | 895 | 396 (44.3) | 0.322 |
| Rheumatoid arthritis, n (%) | 1041 | 21 (2.0) | 880 | 27 (3.1) | 0.148 |
| Glucocorticosteroid use, n (%) | 1041 | 52 (5.0) | 878 | 53 (6.0) | 0.322 |
| Diabetes type 1 or type 2, n (%) | 1041 | 89 (8.6) | 895 | 90 (10.1) | 0.225 |
| History of IBD, n (%) | 1041 | 13 (1.3) | 895 | 7 (0.8) | 0.303 |
| eGFR < 35 ml/min/1.73 m2, n (%) | 1041 | 13 (1.3) | 878 | 18 (2.1) | 0.174 |
| History of stroke, n (%) | 1041 | 82 (7.9) | 895 | 69 (7.7) | 0.891 |
| History of cancer, n (%) | 1041 | 140 (13.5) | 895 | 111 (12.4) | 0.495 |
| Previous AOD use, n (%) | 1041 | 121 (11.6) | 894 | 87 (9.7) | 0.178 |
| Indication for AOD, n (%) | 1041 | 739 (71.0) | |||
| BMD T-score femoral neck | 730 | −1.6 ± 0.9 | |||
| BMD T-score total hip | 731 | −1.1 ± 1.1 | |||
| BMD T-score lumbar spine | 531 | −1.3 ± 1.5 | |||
| Type of index fracture | 1041 | 895 | |||
| Rib(s), sternum, n | 7 | 8 | |||
| Pelvis, n | 34 | 38 | |||
| Upper arm, n | 123 | 143 | 0.210 | ||
| Forearm, n | 319 | 264 | 0.458 | ||
| Wrist, hand, n | 67 | 88 | |||
| Hip, n (%) | 246 | 141 | < 0.001 | ||
| Atypical femur fracture, n | 5 | 0 | |||
| Other fractures of the femur, n | 14 | 11 | |||
| Lower leg, ankle, n | 206 | 173 | |||
| Foot, n | 36 | 53 | |||
| Spine, n | 43 | 26 | 0.144 | ||
| Total number of index fractures, n | 1101 | 933 | |||
| Type of prescribed AOD | |||||
| Alendronate, n (%) | 521 (70.5) | ||||
| Zoledronate (5 mg)b, n (%) | 120 (16.2) | ||||
| Denosumab, n (%) | 79 (8.4) | ||||
| Teriparatide, n (%) | 7 (0.9) | ||||
| Zoledronate (4 mg)c, n (%) | 6 (0.8) | ||||
Data are mean ± SD or number (%)
aVariation in numbers is due to missing. P-values were calculated using t-tests or chi-square
bZoledronate 5 mg was prescribed once a year to treat osteoporosis
cZoledronate 4 mg was prescribed more frequently as adjuvant treatment of patients with breast cancer to prevent bone loss, fracture, and cancer recurrence
FRAX: Fracture Risk Assessment Tool for calculation of the 10-year probability of a major osteoporotic fracture (MOF), IBD: inflammatory bowel disease, eGFR: estimated glomerular filtration rate, AOD: anti-osteoporotic drugs, BMD: bone mineral density. In attenders, 58 patients had two index fractures, and one patient had three index fractures. In non-attenders, 30 patients had two index fractures, and four patients had three index fractures
Patients who were adherent to AOD after 12 months had a higher FRAX-score for MOF (25.7% vs. 23.3%), and a higher percentage used AOD prior to inclusion (16.2% vs. 5.8%) and after 3-month follow-up (86.5% vs. 21.2%) than those who were non-adherent, all p < 0.05 (Table 2).
Table 2.
Characteristics of 427 adherent and 174 non-adherent, Norwegian Capture the Fracture Initiative attenders at the University Hospital of North Norway, Tromsø
| na | Adherent | na | Non-adherent | p | |
|---|---|---|---|---|---|
| Females, n (%) | 427 | 357 (83.6) | 174 | 140 (80.5) | 0.356 |
| Age (years) | 427 | 72.0 ± 9.9 | 174 | 71.1 ± 10.6 | 0.296 |
| Body mass index (kg/cm2) | 380 | 25.2 ± 4.5 | 162 | 25.8 ± 3.9 | 0.144 |
| BMD T-score femoral neck | 261 | −2.1 ± 0.8 | 121 | −2.0 ± 0.6 | 0.621 |
| BMD T-score total hip | 261 | −1.7 ± 0.9 | 121 | −1.6 ± 0.8 | 0.381 |
| BMD T-score lumbar spine | 187 | −2.0 ± 1.3 | 86 | −1.8 ± 1.1 | 0.273 |
| FRAX score, MOF (%) | 349 | 25.7 ± 12.1 | 144 | 23.3 ± 9.1 | 0.016 |
| Currently smoking, n (%) | 411 | 75 (18.3) | 165 | 26 (15.8) | 0.478 |
| High alcohol intake, n (%) | 268 | 18 (6.7) | 98 | 6 (6.1) | 0.840 |
| History of previous fracture, n (%) | 427 | 225 (52.7) | 174 | 81 (46.6) | 0.173 |
| Rheumatoid arthritis, n (%) | 427 | 12 (2.8) | 174 | 2 (1.2) | 0.146 |
| Oral corticosteroid use, n (%) | 427 | 25 (5.9) | 174 | 9 (5.2) | 0.743 |
| Diabetes type 1 or type 2, n (%) | 427 | 33 (7.7) | 174 | 13 (7.5) | 0.915 |
| History of IBD, n (%) | 427 | 6 (1.4) | 174 | 1 (0.6) | 0.306 |
| eGFR < 35 (ml/min/1.73 m2), n (%) | 427 | 1 (0.6) | 174 | 4 (0.9) | 0.625 |
| History of stroke, n (%) | 427 | 40 (9.4) | 174 | 12 (6.9) | 0.302 |
| History of cancer, n (%) | 427 | 59 (13.8) | 174 | 25 (14.4) | 0.860 |
| Previous AOD use, n (%) | 427 | 69 (16.2) | 174 | 10 (5.8) | < 0.001 |
| AOD adherent after 3 months, n (%) | 414 | 358 (86.5) | 170 | 36 (21.2) | < 0.001 |
| Type of prescribed AOD | |||||
| Alendronate, n (%) | 292 (68.4) | 136 (80.0) | |||
| Zoledronate 5 mgb, n (%) | 72 (16.9) | 22 (12.9) | |||
| Denosumab, n (%) | 55 (12.9) | 12 (7.1) | |||
| Teriparatide, n (%) | 5 (1.2) | ||||
| Zoledronate 4 mgc, n (%) | 3 (0.7) | ||||
Data are mean ± SD or number (%)
aVariation in numbers is due to missing. P-values were calculated using t-test or chi-square
bZoledronate 5 mg was prescribed once a year to treat osteoporosis
cZoledronate 4 mg was prescribed more frequently as adjuvant treatment of patients with breast cancer to prevent bone loss, fracture, and cancer recurrence. FRAX = Fracture Risk Assessment Tool for calculation of the 10-year probability of a major osteoporotic fracture (MOF), IBD = inflammatory bowel disease, eGFR = estimated glomerular filtration rate, AOD = anti-osteoporotic drugs, BMD = bone mineral density
During a median 3.5-years observation (range 0.5 to 6.3), attenders had a lower rate of subsequent fracture (57.9 vs. 67.2) and mortality (32.2 vs.49.6) per 1000 person-years compared to non-attenders (Table 3). The attenders to the FLS intervention had 20% lower risk for any type of subsequent fracture (HR 0.80; 95% CI 0.64–0.999) and 33% lower mortality (HR 0.67; 95% CI 0.51–0.89). Those who were adherent to AOD 12 months after initiation of treatment had a lower subsequent fracture rate (53.1 vs. 87.8) per 1000 person-years but not lower mortality than those who were non-adherent. HR for any type of subsequent fracture was 0.51 (95% CI 0.35–0.72) in adherent compared to non-adherent patients. All above-mentioned models were adjusted for age, sex, BMI, smoking, prior fracture, and index hip fracture.
Table 3.
Hazard ratio (HR) with 95% confidence interval (CI) of a subsequent fracture and deaths in attenders vs. non-attenders and adherent vs. non-adherent patients with an index fracture ≥ 50 years at the University Hospital of North Norway, Tromsø 2015–2018
| N | Person- years |
Observation time (years) | Number of fractures | Fracture ratea | Crude HR (95% CI) |
p | HR (95% CI)b |
p | |
|---|---|---|---|---|---|---|---|---|---|
| Non-attenderse | 829 | 2874 | 3.5 | 193 | 67.2 | Reference | Reference | ||
| Attendersc, e | 995 | 3520 | 3.5 | 204 | 57.9 | 0.85 (0.70–1.03) | 0.100 | 0.80 (0.64–0.999) | 0.049 |
| Number of deaths | Mortality ratea | ||||||||
| Non-attenderse | 848 | 3186 | 3.8 | 158 | 49.6 | Reference | Reference | ||
| Attendersc, e | 1025 | 3976 | 3.9 | 128 | 32.2 | 0.66 (0.52–0.783) | < 0.001 | 0.67 (0. 51–0.89) | 0.005 |
| Number of fractures | Fracture ratea | ||||||||
| Non-adherent,e | 168 | 569 | 3.4 | 50 | 87.8 | Reference | Reference | ||
| Adherentd,e | 410 | 1469 | 3.6 | 78 | 653.1 | 0.59 (0.41–0.84) | 0.003 | 0.51 (0.35–0.72) | < 0.001 |
| Number of deaths | Mortality ratea | ||||||||
| Non-adherent | 174 | 674 | 3.9 | 22 | 32.6 | Reference | Reference | ||
| Adherentd | 427 | 1656 | 3.9 | 59 | 35.6 | 1.09 (0.67–1.77) | 0.744 | 1.08 (0.65–1.79) | 0.769 |
aPer 1000 person-year
bAdjusted for age, sex, body mass index, smoking, history of fracture prior to the index fracture, and index hip fracture
cAttenders versus non-attenders to the fracture liaison service intervention
dAdherent versus non-adherent to anti-osteoporotic drugs treatment after 12 months follow-up
eWe excluded the following patients with observation time shorter than six months from the analysis of subsequent fractures: 66 non-attenders, 46 attenders, 6 non-adherent, and 17 adherent patients and from the analyses of mortality: 47 non-attenders, and 16 attenders
Of the 739 patients with fractures and indication for AOD, alendronate was prescribed to 70.5%, zoledronate to 16.2%, denosumab to 8.4% and teriparatide to 0.9% (Table 1). Among attenders with an index hip fracture, alendronate was prescribed to 43.3%, zoledronate to 43.3%, and denosumab to 13.3% of the patients. Alendronate was prescribed to 68.4% of adherent patients, and 80.0% of non-adherent patients (Table 2). The main reasons for non-adherence were disagreement with the diagnosis or indication, gastrointestinal symptoms, dental problems, or unwillingness to take AOD (Table 4).
Table 4.
Reasons of non-adherence to anti-osteoporotic drugs (AOD) treatment at 12-months follow-up was provided by 142 patients
| N (%) | |
|---|---|
| Disagreement with the diagnosis or indication | 30 (21.1) |
| Gastrointestinal symptoms | 28 (19.7) |
| Dental problems | 20 (14.1) |
| Unwillingness to take AOD | 17 (12.0) |
| Fear of rare severe complications (AFF and ONJ) | 11 (7.7) |
| Short life expectancy | 10 (7.0) |
| Musculoskeletal pain | 10 (7.0) |
| Skin rash | 5 (3.5) |
| Other reasons | 11 (7.7) |
*AFF = atypical femur fracture, ONJ = osteonecrosis of the jaw
Among patients with index fractures of the hip, vertebrae, forearm, or upper arm, 95.0%, 79.0%, 83.7% and 87.5% had a BMD T-score < −1.5 at any site (Table 5). The corresponding figures for BMD T-score < −2.5 at any site were 92.5%, 52.6%, 51.7% and 64.6%, and the figures for FRAX score for MOF ≥ 20% were 69.8%, 57.7%, 40.7% and 49.5%, respectively.
Table 5.
Number and proportion (%) of patients with bone mineral density (BMD) T-score at the femoral neck (FN), total hip, lumbar spine and any site < −1.5 and < −2.5 and Fracture Risk Assessment (FRAX) score for a major osteoporotic fracture (MOF) ≥ 20% in 1041 attenders with index fractures at the hip, vertebrae, forearm, and upper arm
| Type of Index fracture: | Hip (n = 247) | Vertebral (n = 36) | Forearm (n = 315) | Upper arm (n = 109) | ||||
|---|---|---|---|---|---|---|---|---|
| n | % | n | % | n | % | n | % | |
| FRAX MOF ≥ 20% | 90/129 | 69.8 | 15/26 | 57.7 | 103/302 | 40.7 | 50/101 | 49.5 |
| FN BMD T-score < −1.5 | 22/25 | 88.0 | 14/19 | 73.7 | 180/285 | 63.2 | 60/90 | 66.7 |
| FN BMD T-score < −2.5 | 17/25 | 68.9 | 3/19 | 15.8 | 38/285 | 13.3 | 19/90 | 21.1 |
| Total Hip BMD T-score < −1.5 | 21/25 | 84.0 | 9/19 | 47.4 | 119/285 | 41.8 | 39/90 | 43.3 |
| Total Hip BMD T-score < −2.5 | 14/25 | 56.8 | 4/19 | 21.1 | 25/285 | 8.8 | 11/90 | 12.2 |
| Spine BMD T-score < −1.5 | 18/31 | 58.1 | 8/14 | 57.1 | 112/212 | 52.8 | 29/56 | 51.8 |
| Spine BMD T-score < −2.5 | 8/31 | 25.8 | 5/14 | 35.7 | 46/212 | 21.7 | 14/56 | 25.0 |
| Any site BMD T-score < −1.5 | 38/40 | 95.0 | 15/19 | 79.0 | 246/294 | 83.7 | 84/96 | 87.5 |
| Any site BMD T-score < −2.5 | 37/40 | 92.5 | 10/19 | 52.6 | 152/294 | 51.7 | 62/96 | 64.6 |
The variation in the denominator was due to a variation in the number who had a BMD-scan
Discussion
In the current study of 1936 fracture patients, 46% were unwilling to participate or did not respond to the invitation. Compared to non-attenders, accepting the FLS intervention was associated with a lower risk of subsequent fractures of any type and mortality during a median of 3.5-years observation. Attenders to the FLS intervention had a 20% lower risk of subsequent fractures and a 33% lower mortality than non-attenders. Patients who attended the FLS and were adherent to AOD after 12 months had a 49% lower risk of subsequent fractures than non-adherent patients.
Several studies have shown that FLS is associated with a 16–40% risk reduction of subsequent fracture and 19–35% reduction in mortality [13–15, 18]. A meta-analysis that excluded studies with high selection bias, showed a 30% reduction in risk of fractures, and a 35% reduction of mortality but this was only found in pre-post FLS comparisons [12]. The current findings agree with these estimates. Figures for the reduction in subsequent fragility fracture risk of 10–13% and mortality of 8–10% in the NoFRACT trial were lower [21]. This is probably a result of a diluting effect of a real-life-setting based on register data. In the NoFRACT trial, patients in the control period (2011–2015) were compared with patients in the FLS-intervention period (2015–2018), without information on whether the patients attended the FLS intervention or received AOD. In contrast, the current study included detailed individual level information about attendance to FLS, adherence to AOD and important confounding factors, such as BMI and smoking. Still, results from this smaller quality assessment study and the larger register-based NoFRACT trial suggest that FLS should be established at departments of orthopedic surgery to improve bone health and reduce mortality. The lower fracture risk in FLS attenders than non-attenders is probably partly explained by differences in health behavior between the groups (healthy selection bias). Nevertheless, the observed results give novel insight into the characteristics and prognosis of patients who choose to reject follow-up treatment, which can be used to inform future intervention strategies to prevent fractures. For example, men who tended to smoke and probably drink more alcohol were less able or willing to attend. When nearly half of the patients were non-attenders, this was partly due to lack of staff or because a written invitation may not be as convincing as a face-to-face invitation. Oral and written information about the possible increased risk of subsequent fracture at the time of treatment for the index fracture through an alliance between the FLS-nurses, orthopedic surgeons, and other health care workers involved in the fracture treatment may improve attendance. In addition, we need to continuously share knowledge on the importance of osteoporosis prevention and treatment, including FLS, with politicians, hospital leaders, physicians, nurses, general practitioners, patients, spouses, and the general population.
In the current study, 58% of patients who had AOD prescribed were adherent to this treatment 12 months after initiation. This percentage is higher than previously reported in studies using the Norwegian prescription database, where only 15% of women and 4% of men with hip fractures, and 11% of women and 3% of men with forearm fractures used AOD the first year after the fracture [24, 25]. The implementation of the current FLS intervention clearly increased treatment initiation. The high rates of prescription in the current study may be due to the omission of BMD scanning in high-risk patients and using BMD < −1.5 as the treatment threshold. The nurse spent a lot of time educating and monitoring patients’ treatment during the assessment of their fracture risk and during the follow-up phone calls. In the case of side effects, an alternative treatment was suggested. This might have contributed to an increased understanding of the importance of osteoporosis treatment and improved motivation and adherence. Patients’ involvement is reported to be important for treatment adherence [26, 27].
Of the non-adherent patients, 80% were prescribed oral treatment. Of the adherent patients, 68% were prescribed oral treatment, and over 32% used intravenous or subcutaneous treatment. Not being adherent to AOD after 3 months seems to be a major factor for not being adherent after 12 months. Some of these patients did not initiate AOD and some discontinued early. The main reasons for non-adherence to AOD after the 12 months follow-up were disagreement with the osteoporosis-diagnosis or indication for treatment, gastrointestinal side effects, dental problems, and unwillingness to take AOD. Some patients disagreed that they had osteoporosis, some believed that they did not need treatment, while others were skeptical to the benefit of the treatment, as reported before [28]. After 12 months, some of the patients’ general practitioners or physicians at the nursing homes suggested that patients had become too frail or had too short life expectancy for further treatment.
In NoFRACT, a hip fracture was considered synonymous with osteoporosis. Many hip fracture patients were therefore treated without a BMD scan for assessment of the diagnosis. Only 16% of hip fracture patients had a BMD scan, and 95.0% and 92.5% of them had a BMD T-score < −1.5 and < −2.5, respectively. This justifies that AOD can be prescribed without prior BMD measurements to hip fracture patients. For patients with other types of fractures, BMD was more important for the diagnosis. Moreover, we used BMD T-score < −1.5 as a treatment threshold because the commonly used T-score < −2.5 may be too restrictive as most of the patients with fracture have BMD T-score > −2.5 [29].
To bridge the treatment gap and ensure that patients with osteoporotic fractures receive AOD when needed, it is recommended to establish an FLS led by a dedicated coordinator at every hospital treating fracture patients [20, 21]. In this recommendation, patients suffering a hip, or a vertebral fracture should be offered AOD regardless of BMD T-score and/or FRAX score, to reduce treatment delay. There are several barriers to the implementation of a successful FLS program, such as lack of knowledge about osteoporosis and the effectiveness of AOD in fracture prevention [30]. Only 8% of the patients in this study reported that they never started or stopped treatment in fear of rare side effects such as osteonecrosis of the jaw. Serious adverse side effects are rare compared to the higher incidence of subsequent fracture following an initial fracture [30].
The limitations of this study are the moderate sample size and 20% was lost to follow-up, thus findings should be interpreted with caution. Some patients were unable or unwilling to attend FLS, and sometimes there was a lack of capacity by the staff to offer FLS to the eligible patients, thus only 53.8% attended. Selection is likely in our data, as non-attenders who were unable or unwilling to receive the FLS intervention, were to a larger extent men, who were older and smoked more than the attenders. It is unfortunately not possible to distinguish the effect of the intervention itself from the effect of healthy selection within our dataset. Although we started the observation window six months after the time of the index fracture, the results may, at least partly, reflect the effect of being able and willing to participate (healthier, more well-educated individuals attend and adhere to AOD). In addition, information on education is missing and may be a relevant confounder. The proportion of patients who were invited by letter is not registered. However, no clear difference in prevalence of diseases between non-attenders and attenders was identified. Moreover, the data on fracture cases may be incomplete as fractures occurring during travels might be lost from registration. However, this is not likely to represent an important problem in this age group. Lack of mortality benefit of AOD adherence may be from the small sample size.
Strengths of this study include that all patients had their fractures radiologically confirmed by a radiologist or an orthopedic surgeon. The medical records, including the descriptions of the fracture-locations, were thoroughly reviewed for all attenders and non-attenders. At follow-up, the nurses collected detailed information on adherence to AOD from the patients, their spouses, home nurses, or nurses at the nursing home in cases where the patients were unable to respond.
In conclusion, this quality assessment study showed reduced subsequent fracture rates and mortality in patients who were able and willing to attend FLS. Patients who were adherent to AOD had a reduced risk of subsequent fracture but the same mortality risk. There is still a care gap between patients who need secondary fracture prevention and those who receive this. Increased focus on the importance of osteoporosis and fracture prevention, as well as patient involvement, may improve attendance and adherence rates in the future.
Acknowledgements
We thank all personnel at the participating site for their work in FLS; Marit Osima, May-Greta Pedersen, and Anita Kanniainen (University Hospital of North Norway).
Author contributions
ÅB, TTB, FF, TKO, and LBS designed the study, secured the fundings and were on the study steering committee. ÅB performed the study analysis. ÅB and JK have directly accessed and verified the underlying data reported in the manuscript. ÅB, and JK drafted the manuscript that was written in collaboration with all co-authors. All authors approved the final version of the manuscript. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.
Funding
Open access funding provided by UiT The Arctic University of Norway (incl University Hospital of North Norway) The study was supported with grants from; The Regional Health Authorities, grant number 243852; Northern Norway Regional Health Authority, grant number 14083. The funders of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report.
Data availability
Due to protection of privacy under General Data Protection Regulation and Norwegian law, the individual-level data can only be made available after approval by the Data Protection Officer at the University Hospital of North Norway.
Declarations
Conflict of interest
ÅB, JK, AKH, TKO and FIN, no support from any organization for the submitted work; FF reports lecturing fees from UCB and Amgen; TTB reports speaker fees from UCB, Amgen, Roche Diagnostics, and Pharma Prim and participation in an advisory board for UCB, LBS reports speaker fees from Eli Lilly, Amgen and UCB.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Ahmed LA, Schirmer H, Bjornerem A, Emaus N, Jorgensen L, Stormer J et al (2009) The gender- and age-specific 10-year and lifetime absolute fracture risk in Tromso. Norway Eur J Epidemiol 24(8):441–448 [DOI] [PubMed] [Google Scholar]
- 2.Omsland TK, Emaus N, Tell GS, Magnus JH, Ahmed LA, Holvik K et al (2014) Mortality following the first hip fracture in Norwegian women and men (1999–2008). A NOREPOS study. Bone 63:81–86 [DOI] [PubMed] [Google Scholar]
- 3.Lotters FJ, van den Bergh JP, de Vries F, Rutten-van Molken MP (2016) Current and future incidence and costs of osteoporosis-related fractures in the Netherlands: combining claims data with BMD measurements. Calcif Tissue Int 98(3):235–243 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chandran M, Akesson KE, Javaid MK, Harvey N, Blank RD, Brandi ML et al (2024) Impact of osteoporosis and osteoporosis medications on fracture healing: a narrative review. Osteoporos Int 35(8):1337–1358 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Omsland TK, Holvik K, Meyer HE, Center JR, Emaus N, Tell GS et al (2012) Hip fractures in Norway 1999–2008: time trends in total incidence and second hip fracture rates. A NOREPOS study. Eur J Epidemiol 27(10):807–814 [DOI] [PubMed] [Google Scholar]
- 6.Lyles KW, Colon-Emeric CS, Magaziner JS, Adachi JD, Pieper CF, Mautalen C et al (2007) Zoledronic acid and clinical fractures and mortality after hip fracture. N Engl J Med 357(18):1799–1809 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cosman F, de Beur SJ, LeBoff MS, Lewiecki EM, Tanner B, Randall S et al (2014) Clinician’s guide to prevention and treatment of osteoporosis. Osteoporos Int 25(10):2359–2381 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Langdahl BL, Silverman S, Fujiwara S, Saag K, Napoli N, Soen S et al (2018) Real-world effectiveness of teriparatide on fracture reduction in patients with osteoporosis and comorbidities or risk factors for fractures: integrated analysis of 4 prospective observational studies. Bone 116:58–66 [DOI] [PubMed] [Google Scholar]
- 9.Lane J, Langdahl B, Stone M, Kurth A, Oates M, Timoshanko J et al (2024) Romosozumab in patients who experienced an on-study fracture: post hoc analyses of the FRAME and ARCH phase 3 trials. Osteoporos Int 35(7):1195–1204 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tsourdi E, Langdahl B, Cohen-Solal M, Aubry-Rozier B, Eriksen EF, Guanabens N et al (2017) Discontinuation of denosumab therapy for osteoporosis: a systematic review and position statement by ECTS. Bone 105:11–17 [DOI] [PubMed] [Google Scholar]
- 11.Briot K (2017) Fracture liaison services. Curr Opin Rheumatol 29(4):416–421 [DOI] [PubMed] [Google Scholar]
- 12.Li N, Hiligsmann M, Boonen A, van Oostwaard MM, de Bot R, Wyers CE et al (2021) The impact of fracture liaison services on subsequent fractures and mortality: a systematic literature review and meta-analysis. Osteoporos Int 32(8):1517–1530 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Huntjens KM, van Geel TA, van den Bergh JP, van Helden S, Willems P, Winkens B et al (2014) Fracture liaison service: impact on subsequent nonvertebral fracture incidence and mortality. J Bone Joint Surg Am 96(4):e29 [DOI] [PubMed] [Google Scholar]
- 14.Nakayama A, Major G, Holliday E, Attia J, Bogduk N (2016) Evidence of effectiveness of a fracture liaison service to reduce the re-fracture rate. Osteoporos Int 27(3):873–879 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Axelsson KF, Jacobsson R, Lund D, Lorentzon M (2016) Effectiveness of a minimal resource fracture liaison service. Osteoporos Int 27(11):3165–3175 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hawley S, Javaid MK, Prieto-Alhambra D, Lippett J, Sheard S, Arden NK et al (2016) Clinical effectiveness of orthogeriatric and fracture liaison service models of care for hip fracture patients: population-based longitudinal study. Age Ageing 45(2):236–242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.de Bruin IJA, Wyers CE, van den Bergh JPW, Geusens P (2017) Fracture liaison services: do they reduce fracture rates? Ther Adv Musculoskelet Dis 9(7):157–164 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.van Geel T, Bliuc D, Geusens PPM, Center JR, Dinant GJ, Tran T et al (2018) Reduced mortality and subsequent fracture risk associated with oral bisphosphonate recommendation in a fracture liaison service setting: A prospective cohort study. PLoS ONE 13(6):e0198006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Axelsson KF, Johansson H, Lundh D, Moller M, Lorentzon M (2020) Association between recurrent fracture risk and implementation of fracture liaison services in four Swedish hospitals: a cohort study. J Bone Miner Res 35(7):1216–1223 [DOI] [PubMed] [Google Scholar]
- 20.Andreasen C, Solberg LB, Basso T, Borgen TT, Dahl C, Wisloff T et al (2018) Effect of a fracture liaison service on the rate of subsequent fracture among patients with a fragility fracture in the Norwegian capture the fracture initiative (NoFRACT): a trial protocol. JAMA Netw Open 1(8):e185701 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Andreasen C, Dahl C, Frihagen F, Borgen TT, Basso T, Gjertsen JE et al (2025) Fracture liaison service (FLS) is associated with lower subsequent fragility fracture risk and mortality: NoFRACT (the Norwegian capture the fracture initiative). Osteoporos Int 36(3):501–512 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wu CH, Chen CH, Chen PH, Yang JJ, Chang PC, Huang TC et al (2018) Identifying characteristics of an effective fracture liaison service: systematic literature review. Osteoporos Int 29(5):1023–1047 [DOI] [PubMed] [Google Scholar]
- 23.Kanis JA, Cooper C, Rizzoli R, Reginster JY (2019) European guidance for the diagnosis and management of osteoporosis in postmenopausal women. Osteoporos Int 30(1):3–44 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hoff M, Skurtveit S, Meyer HE, Langhammer A, Søgaard AJ, Syversen U et al (2015) Use of anti-osteoporotic drugs in central Norway after a forearm fracture. Arch Osteoporos 10(1):30 [DOI] [PubMed] [Google Scholar]
- 25.Devold H, Søgaard A, Tverdal A, Falch J, Furu K, Meyer H (2013) Hip fracture and other predictors of anti-osteoporosis drug use in Norway. Osteoporos Int 24(4):1225–1233 [DOI] [PubMed] [Google Scholar]
- 26.Hiligsmann M, Cornelissen D, Vrijens B, Abrahamsen B, Al-Daghri N, Biver E et al (2019) Determinants, consequences and potential solutions to poor adherence to anti-osteoporosis treatment: results of an expert group meeting organized by the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO) and the International Osteoporosis Foundation (IOF). Osteoporos Int 30(11):2155–2165 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mangano GRA, Avola M, Blatti C, Caldaci A, Sapienza M, Chiaramonte R et al (2022) Non-adherence to anti-osteoporosis medication: factors influencing and strategies to overcome it. A narrative review. J Clin Med 12(1):14 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Silverman SL, Schousboe JT, Gold DT (2011) Oral bisphosphonate compliance and persistence: a matter of choice? Osteoporos Int 22(1):21–26 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Siris ES, Chen YT, Abbott TA, Barrett-Connor E, Miller PD, Wehren LE et al (2004) Bone mineral density thresholds for pharmacological intervention to prevent fractures. Arch Intern Med 164(10):1108–1112 [DOI] [PubMed] [Google Scholar]
- 30.Black DM, Geiger EJ, Eastell R, Vittinghoff E, Li BH, Ryan DS et al (2020) Atypical femur fracture risk versus fragility fracture prevention with bisphosphonates. N Engl J Med 383(8):743–753 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Due to protection of privacy under General Data Protection Regulation and Norwegian law, the individual-level data can only be made available after approval by the Data Protection Officer at the University Hospital of North Norway.