Artikel: Screening for osteoporosis 2017

(1)

ORIGINAL ARTICLE

Screening for osteoporosis following non-vertebral fractures in patients aged 50 and older independently of gender or level of trauma energy — a Swiss trauma center approach

Christoph Hemmeler¹_&Sabrina Morell²_&Felix Amsler³_&Thomas Gross²

Received: 27 December 2016 / Accepted: 3 April 2017

#International Osteoporosis Foundation and National Osteoporosis Foundation 2017

Abstract

Summary Screening in a standardized manner for osteoporosis in non-vertebral fracture patients aged 50 and older independently of both gender and level of trauma energy yielded the indication for osteoporotic therapy for every fourth male high-energy fracture patient.

PurposeThis study aimed to identify the rate of osteoporosis in patients of both genders after fracture independently of the underlying level of trauma energy.

Methods A random cohort of patients aged 50 or older with non-vertebral fractures participated in a standardized diagnostic protocol to evaluate the indication for treatment of osteoporosis (number needed to screen (NNS)). Univariate and multivariate analysis as well as correlation testing were performed to determine statistical relationships. Significance was set atp< 0.05.

Results Of 478 fracture patients with a mean age of 69.3 ± 11.8 years, 317 (66.3%) were female and 161 (33.7%) male. One hundred nineteen patients (24.9%) sustained high-energy fractures (HEFs) and 359 (75.1%) low-energy fractures (LEFs). Twenty-eight percent of males and 47% of females qualified as osteoporotic in densitometry

(dual-energy X-ray absorptiometry (DXA)), resulting in a NNS of 2.1 for women and 3.6 for men. The indication for treatment of osteoporosis increased to an NNS of 1.5 for females and 2.4 for males if the fracture risk assessment tool (FRAX) was included in the diagnostics (DXA and FRAX).

With regard to the energy of trauma, the NNS for treatment following DXA and FRAX was 1.5 for LEF and 2.9 for HEF.

Subgroup analysis revealed that HEF males within the decennia 50+ and 80+ had an NNS of around 3, i.e., comparable to females and about twice as high as LEF patients.

Conclusions These preliminary findings appear to confirm the pragmatic approach to screening in a standardized manner for osteoporosis in all non-vertebral fracture patients aged 50 and older—independently of both gender and level of trauma energy.

Keywords Osteoporosis . Fracture . Densitometry . Fracture risk assessment tool . Gender . Trauma energy

Introduction

The importance of adequate diagnostics and osteoporosis therapy in the context of fracture treatment is being increasingly acknowledged not only in the literature but also in daily orthopedic practice [1,2]. Osteoporosis results in an increased risk of fragility fractures [3]. These fractures are associated with substantial pain and suffering and frequently require operative stabilization. Despite all efforts, they often result in disability, increased mortality, and substantial costs to society [4]. In 2014, 74,000 cases in Switzerland cost an estimated 1558 million Swiss Francs [5]. In the USA, two million fragility fractures occurred in 2005 with costs of $17 billion. This is expected to rise by 50% by 2025 [6]. Following an osteoporotic fracture and depending on age, the risk of a subsequent Electronic supplementary materialThe online version of this article

(doi:10.1007/s11657-017-0334-3) contains supplementary material, which is available to authorized users.

* Thomas Gross thomas.gross@ksa.ch

1 Department of Rheumatology, Cantonal Hospital, Tellstr.1, CH-5001 Aarau, Switzerland

2 Department of Traumatology, Cantonal Hospital, Trauma Unit, Tellstr.1, CH-5001 Aarau, Switzerland

3 Amsler Consulting, Gundeldingerrain 111, CH-4059 Basel, Switzerland

(2)

fracture varies between 9 and 30% in women and 10 and 26%

in men [6,7].

With regard to osteoporotic fractures, most studies and guidelines focus on reduced bone mineral density in females following low-energy trauma (measured by dual-energy X- ray absorptiometry (DXA)) as an indicator of a patient’s subsequent risk of further fractures.

The implicit assumption is that fractures in male patients or those resulting from high-energy trauma (HEF) do not have an elevated fracture risk [8,9]. Consequently, in clinical practice, male and/or HEF patients rarely undergo osteoporotic diagnostics with the potential consequence of insufficient secondary prevention and misinterpretation of the socioeconomic impact of osteoporotic fractures [4,5].

On the other hand, recent data indicate reduced bone mineral density and an increased rate of osteoporosis even in older men and following HEF. For example, the Geelong Osteoporosis Study demonstrated that women over 50 years who sustained HEF had significantly reduced bone mineral density compared to a control cohort of the population [9].

Mackey found a positive correlation with subsequent fractures for both HEF and LEF in both women and men over 65 [8].

As a consequence, even though almost all studies on the topic focused on elderly women and/or specific fracture regions, there is evidence to suggest that diagnostics for osteoporosis might be very worthwhile after high-energy fractures [8, 9]

and for males [10]. To exacerbate the situation, the exact def- initions of HEF and LEF differ between studies, if cited at all [11,12]. Further, the age threshold for recommendation for osteoporosis diagnostics differs significantly between national and specialty guidelines, e.g., following menopause or starting at 50 or 65 years of age depending also on gender [13–15].

With this in mind and as a teaching hospital, we decided to screen all fracture patients aged 50 or older for osteoporosis as part of daily clinical routine, independently of gender and trauma energy, and to recommend therapy if indicated. With this simple cutoff, we tried to keep the organizational barrier for diagnostics low. On the other hand, we were interested in investigating this apparently aggressive screening for osteoporosis in detail with the aim of finding out which patients and fracture constellations would profit from a standardized procedure.

Therefore, the primary objective of this investigation in patients 50 years and older who had sustained non-vertebral fractures was to identify the effective rate of cases in our department in which osteoporotic therapy would be required.

The resulting number needed to screen (NNS) was differenti- ated for the specific risk groups: male versus female, HEF vs.

LEF, and between stratified age groups (decennia). Given the finding that many professionals and laymen still only use pathologic osteodensitometry (DXA) values to define osteoporosis and identify the need to initiate osteoporotic therapy, we additionally aimed to discover how many more cases in

our cohort received therapy if fracture risk assessment tool (FRAX) scoring was performed in addition to DXA.

Methods

Between March 2012 and November 2014, patients of both genders aged 50 or older and hospitalized in the trauma unit of the hospital for non-vertebral fractures were prospectively investigated for osteoporosis following a standardized protocol approved by the Cantonal Ethics Board (EK2013/036;

NCT02157753). The teaching hospital (1 of 12 Swiss highly specialized major trauma centers) treats >800 fracture patients a year on a non-ambulatory basis at its trauma unit. Of those, about 60% are aged 50 or older, 70% of the latter being oper- ated on, and all of them subject to the described procedure.

Bone mineral density (BMD) was measured using a standardized procedure with DXA for the above patients as part of the outpatient procedure (S/N 82833, Hologic Inc., Discovery W, Waltham, MA, USA). The measurements were performed by specifically trained and certified operators at the proximal femur (neck and total), the lumbar spine, and the distal third of the radius. Physicians interpreting DXA scans were fully certified and had extensive experience in this area. The lowest value obtained was used for the analysis.

Patients were categorized as having osteopenia (a DXAT score of −1.1 to −2.4) or osteoporosis (a DXA Tscore of

≤−2.5) [16]. Those with aTscore≤-2.5 and hip or vertebral fracture without adequate trauma were directly considered for treatment.

For cases of osteopenia, FRAX scoring was additionally performed drawing upon the guidelines of the Swiss Association Against Osteoporosis (SVGO) 2010 and based on DXA and FRAX; the absolute risk of fracture in each age decennia was then estimated [16]. As a reference, the latest National Osteoporosis Foundation (NOF) guidelines 2014 give a low bone mass (Tscore between−1.0 and−2.5 at the femoral neck or lumbar spine) and a 10-year probability of a hip fracture ≥3% or a 10-year probability of a major osteoporosis-related fracture≥20% [15].

In addition to the analysis of basic medical records, a standard questionnaire to record risk factors for osteoporosis was distributed to enable further evaluation using the World Health Organization (WHO) online FRAX. We used the algorithm for Switzerland to estimate the 10-year fracture risk for indi- vidual patients. Registered risk factors of osteoporosis were positive family history of femoral neck fracture, smoking, current intake of alcohol, cortisone or aromatase inhibitors, low body mass index (<20 kg/m²), decrease of body height (>3 cm), COPD, IBD, signs of hypogonadism, or rheumatoid arthritis. On the basis of these outcomes for each option, i.e., DXA only or DXA combined with FRAX, the patients were stratified into those having osteoporosis (yes/no) and those

(3)

qualifying for treatment (treatment indication yes/no).

Furthermore, an extensive laboratory workup was performed including values for liver and kidney function, erythrocyte sedimentation rate, calcium, 25-hydroxy vitamin D, phos- phate, albumin, thyroid-stimulating hormone (TSH), parathy- roid hormone (PTH), as well as a hemogram. Evaluation of these measurements, including FRAX, was conducted in a standardized manner by a specialist from the Department of Rheumatology (C.H.). A subsequent therapy recommendation was forwarded to the family doctor and the orthopedic sur- geon in charge for each patient.

Data for this investigation were collected by a trained study nurse (S.M.), specifically evaluating parameters such as low- energy vs. high-energy trauma or the age unadjusted and ad- justed Charlson scores [17,18] to determine patients’comor- bidities in a standardized manner. Missing data were obtained by contacting patients and/or primary care physicians. With regard to the energy of trauma, LEF was defined as falls from standing height or less or stairs≤50 cm high or injuries resulting from walking velocity or less. In contrast, HEF was defined as falls from greater than standing height or stairs more than 50 cm high or with a velocity >5 km/h (modified accordingly [8,9]). Standard stratification was undertaken by the study nurse based on patients’or independent observers’

descriptions of the relevant trauma mechanism as evaluated in the emergency department at the time of hospital admission.

With regard to age, the decennials 50–59 (50+), 60–69 (60+), 70–79 (70+), and 80 years of age and older (80+) were investigated separately. NNS to find a patient with osteoporosis confirmed by densitometry (DXA) or with the indication for treatment following additional FRAX (DXA and FRAX) was calculated and presented with 95% confidence intervals.

Statistical analysis

Chi-squared tests were used to compare binary variables and analysis of variance (ANOVA) to compare mean values of continuous variables. Correlations between categorized osteoporosis diagnosis and laboratory values were calculated using non-parametric Spearman correlations. Forward stepwise multivariate logistic regression analysis was performed to identify independent predictors of the indication for treatment of osteoporosis. The significance of each variable was assessed using the likelihood ratio test. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. Allp values are two-tailed.

Results

During the study period, 849 patients qualified for inclusion.

Three hundred seventy-one patients (43.7%) did not undergo the standard procedure. The reasons for excluding patients

were not available for every single case, but major reasons were as follows: The stipulated process was not followed through; screening for osteoporosis was performed at a different institution, mostly with DXA only; participating patients, their next of kin, care, and or nursing homes or doctors did not agree to the procedure. Detailed univariate and multivariate comparison of the study cohort with the group of excluded patients is given in supplemental Tables Aand B. Overall, excluded patients more often lived in a nursing home (i.e., not in their own homes) and more often underwent conservative therapy of their fracture. No significant differences were found between groups with regard to gender, length of hospital stay, or inclusion rate over the study period.

A total of 478 patients aged 50 or older with non-vertebral fractures hospitalized at the trauma unit of the Cantonal Hospital Aarau were investigated for osteoporosis.

Characteristics of the study cohort are given in Table1.

Two hundred two fractures were localized in the lower extremities, 203 in the upper extremities; all other cases involved the torso or multiple fractures.

At the time of injury, 5.2% of cases were known to have osteoporosis.

The mean interval between the first clinical visit after fracture and subsequent DXA performed according to the protocol was 71 ± 40 days.

Overall, 13.8% of examined patients revealed normal T score values (Table2). In comparison to a DXA osteoporosis rate of 40.6%, the indication for treatment increased to 59.4%

if FRAX was included. The resulting NNS to identify indica- tions for osteoporotic treatment was 2.5 for DXA only compared to 1.7 for the combined use of DXA and FRAX.

In univariate analysis in relation to gender, both men and women presented with normalTvalues (19.3% [n= 31] and 11.0% [n= 35], respectively) in less than one fifth of cases.

Given a total rate of 28% of male and 47% of female fractures qualifying as osteoporotic in DXA, the resulting NNS was 2.1 for women and 3.6 for men. The indication for treatment of osteoporosis rose to 41% in male and 69% in female fracture patients if FRAX screening was included. The overall NNS for recommendation of treatment based on combined DXA and FRAX risk assessments was 1.5 for women and 2.4 for men.

With regard to the energy of injury, 24.9% of patients suf- fered high-energy trauma. These HEF patients were more often male, younger, healthier (Charlson), and presented more often with multiple fractures and a shorter length of stay than low-energy trauma patients (allp< 0.001; Table1).

Operative fixation rate was comparable between groups.

Detailed analysis revealed significantly more multiple fractures and, at the same time, less radial and femoral fractures in high-energy trauma patients (Table2). In both LEF (10.0%) and HEF (25.2%), only a minority of patients demonstrated normalTscore values (Table2).

(4)

Given a rate of 46.2% LEF and 23.5% HEF presenting as osteoporotic in DXA, the resulting NNS was 2.16 for LEF and 4.25 for HEF, respectively. The indication for treatment of osteoporosis rose to 67.6% in LEF and 34.4% in HEF if FRAX screening was included. The overall NNS for recommendation of treatment based on combined DEXA and FRAX was 1.5 for LEF and 2.9 for HEF.

Combined analysis, which allowed for the influences of gender and trauma energy, revealed that male HEF patients within the decennia 50+ had an NNS of 3 (13 patients out of 39) and those who were 80+ of 2 (3 patients out of 6) for the treatment of osteoporosis (DXA and FRAX), i.e., similar to female patients and about twice as high as LEF patients.

Figures 1 and2 illustrate these findings in comparison to age, gender, and trauma energy specific prevalence data for osteoporosis from Sweden [21].

In contrast, male HEF patients of the decennia 60+ to 70+

had a conspicuously much lower risk in terms of treatment needs (NNS 15 (1 patient out of 15) and 11 (1 patient out of 11), respectively).

Overall, increasing age was associated with a higher rate of osteoporosis (p= 0.018). High significance (p< 0.001) was found for both, for DXA stratification only (mean 63.5 ± 9.5 years normal, 66.1 ± 10.2 years osteopenic, 74.8 ± 11.8 years osteoporotic) and for the combined assessment to indicate treatment using DXA and FRAX (treatment indicated 70.9 ± 12.2 years/treatment not indicated 66.9 ± 10.7 years). Stratified for decennial age groups beginning at the age of 50 years (Table3), it was found that below the age of 80, the NNS for the treatment of osteoporosis (DXA and FRAX) in high-energy fracture patients was about twice as high as for the low-energy group (e.g., for patients 50+

years of age 3.1 in HEF vs. 1.4 in LEF).

In patients of 80 years and older, this difference related to trauma energy almost disappeared (Fig.2). Including gender, the overall NNS of 1.5 for the treatment of osteoporosis (DXA and FRAX) in female patients (Table3) remained almost un- changed over all decennia in contrast to male patients overall NNS of 2.4 demonstrating a peak of 4.8 at the decennium 70+

and arriving at the NNS for women at 80+ (Fig.2).

Detailed univariate analysis of lab values or specific risk constellations such as hyperparathyroidism or smoking did not reveal any indicative correlations with the indication for the need to treat osteoporosis (all rho < 0.2; data not shown).

In multivariate analysis using a model that included the variables identified as significant by univariate analysis, namely LEF, female gender, and the fracture regions femur, pelvis, and radius, they were found to account for 23.3% of variance for the need of osteoporotic therapy (Table4). In a logistic regression model including only age, gender, and trauma energy, age was not significant and female gender together with low-energy trauma explained 15.9% of the variance for the need for osteoporotic therapy. In this model, no interaction Table1Characteristicsofthestudycohort(n=478) TotalSexTrauma Male(N=161)Female(N=317)pLowenergy(N=359)Highenergy(N=119)p Genderfemale317(66.3%)–––269(74.9%)48(40.3%)<0.001 Gendermale161(33.7%)––90(25.1%)71(59.7%) Low-energytrauma359(75.1%)90(55.9%)269(84.9%)<0.001––– High-energytrauma119(24.9%)71(44.1%)48(15.1%)–– Age(attimeofDXA)69.3±11.865.7±11.271.1±11.6<0.00171.2±11.663.7±10.3<0.001 AgeunadjustedCharlsonscore0.68±1.230.63±1.110.71±1.290.4890.8±1.310.32±0.82<0.001 AgeadjustedCharlsonscore3.79±1.683.47±1.593.95±1.710.0034.03±1.733.04±1.27<0.001 Operativefracturefixation423(88.5%)142(88.2%)281(88.6%)0.886320(89.1%)103(86.6%)0.445 Hospitallengthofstay9.9±7.811.4±8.49.2±7.30.0049.1±6.612.5±10.2<0.001 DXAosteoporosis194(40.6%)45(28.0%)149(47.0%)<0.001166(46.2%)28(23.5%)<0.001 IndicationfortherapyDXAandFRAX284(59.4%)66(41.0%)218(68.8%)<0.001243(67.7%)41(34.5%)<0.001 DXAosteodensitometry,FRAXfractureriskassessmenttool

(5)

effect could be found between gender and trauma energy (p= 0.545).

Discussion

To our knowledge, this prospective investigation is the first clinical evaluation with the specific objective of systematically screening fracture patients aged 50 or older in a simple manner for osteoporosis, independently of gender and the level of trauma energy involved. Subsequently, therapy was rec- ommended to patients if indicated based on a standard protocol. The primary objective was to identify the actual rate of cases (NNS) requiring osteoporotic therapy following non- vertebral fractures. The secondary objective was to further stratify the resulting NNS for the major risk groups of male versus female, HEF vs. LEF, and age decennia.

First, the overall results are in line with those published in more recent literature to show that female gender, low-energy fracture, and specific fracture regions (such as the femur) are the main indicators of osteoporosis [19].

Second, we found an impressively low overall NNS to detect patients needing osteoporotic therapy, independently of gender or trauma energy and in all age groups over 50.

Even the subgroup demonstrating the lowest efficacy in the investigation, i.e., HEF male patients aged 60–70 yielded an NNS of 15. Of course, one can argue that for 60+ and 70+

male, this NNS corresponds to known osteoporosis prevalence data such as those from Sweden [21]. But for all other groups under investigation, demonstrated NNS were consid- erably lower than expected from prevalence data. In addition, such numbers have to be viewed in the context of the fact that, for example, for patients taking statins as the most widely accepted standard prevention after stroke, there is an NNS

>30 [20]. Against this background, it appears worthwhile to systematically screen fracture patients aged 50 or older for osteoporosis, independently of gender and the level of trauma energy involved.

Third, this study revealed an indication for osteoporosis treatment in a surprisingly high number of HEF cases: The corresponding NNS following HEF was 2.9 vs. 1.5 in LEF.

Fig. 2 Relationship of NNS and age in comparison to level of trauma energy and population of males and females for osteoporosis from Sweden

Fig. 1 Relationship of NNS and age in comparison to gender, level of trauma energy, and population for osteoporosis from Sweden

Table 2 DXA and FRAX results according to gender and trauma energy (N= 478) Trauma energy Gender NormalT

score≥−1

OsteopeniaTscore−1.1 to−2.5 OsteoporosisT score <−2.5

Total DXA alone DXA and FRAX

No indication for treatment (FRAX)

Indication for treatment (FRAX)

NNS 95% CI NNS 95% CI

Low Male 10 (11.1%) 32 (35.6%) 17 (18.9%) 31 (34.4%) 90 (100%) 2.90 2.26–4.06 1.88 1.57–2.32 Female 26 (9.7%) 48 (17.8%) 60 (22.3%) 135 (50.2%) 269 (100%) 1.99 1.78–2.26 1.38 1.28–1.49 Total 36 (10%) 80 (22.3%) 77 (21.4%) 166 (46.2%) 359 (100%) 2.16 1.95–2.43 1.48 1.38–1.59 High Male 21 (29.6%) 32 (45.1%) 4 (5.6%) 14 (19.7%) 71 (100%) 5.07 3.45–9.56 3.94 2.82–6.56 Female 9 (18.8%) 16 (33.3%) 9 (18.8%) 14 (29.2%) 48 (100%) 3.43 2.38–6.13 2.09 1.61–2.96 Total 30 (25.2%) 48 (40.3%) 13 (10.9%) 28 (23.5%) 119 (100%) 4.25 3.21–6.29 2.90 2.33–3.86 Total Male 31 (19.3%) 64 (39.8%) 21 (13%) 45 (28%) 161 (100%) 3.58 2.87–4.76 2.44 2.06–2.99 Female 35 (11%) 64 (20.2%) 69 (21.8%) 149 (47%) 317 (100%) 2.13 1.90–2.41 1.45 1.35–1.57 Total 66 (13.8%) 128 (26.8%) 90 (18.8%) 194 (40.6%) 478 (100%) 2.46 2.22–2.76 1.68 1.57–1.82 NNSnumber needed to screen,DXAosteodensitometry,FRAXfracture risk assessment tool

(6)

With the exception of a few high-risk constellations, many authors and guidelines appear to keep the approach to osteoporotic fractures very limited by making the diagnosis dependent on low-energy trauma only [5, 21]. In this way, only patients without a so-called adequate trauma [22] are investigated for osteoporosis. This practice seems astonishing given the lack of evidence as to where the exact line is between low- energy and high-energy trauma [23,24]. Additionally, studies have questioned the assumption that a healthy bone will not break on low-energy impact [25,26]. In fact, even a fall from standing height may generate enough energy to fracture a normal bone [27]. There is good evidence that both low- trauma and high-trauma fractures are strongly associated with low BMD and an increased risk for subsequent fractures in older women and men and should therefore be included in observational studies [8]. Several authors have suggested that any prior fracture whether a high-trauma or low-trauma one may indicate underlying skeletal fragility and an increased risk for future fracture [9, 19]. The authors of a meta- analysis including randomized fracture prevention trials

summarized that the potential benefits of treatment for an in- dividual patient would be best described for all non-vertebral fractures, not just a subset such as low-trauma fragility fractures [28]. Against this background, it is our opinion that our prospective cohort data on non-vertebral fracture patients aged 50 or older argue for an aggressive screening approach for osteoporosis, independently of the level of trauma energy involved. In spite of these strong arguments, the qualifying statement must be made that to date, evidence on the effectiveness of antiosteoporosis treatment in patients with high- trauma fractures is missing.

Fourth, the observed difference in NNS between men and women of the same age or fracture type was less than gener- ally expected; for example, the NNS was about 5 in 70+ year- old male in comparison to a NNS of about 1.5 in 70+ female.

The risk for osteoporosis in males continues to be underestimated by both the population and by medical doctors with the consequence that male fracture patients less common- ly undergo any diagnostic tests for osteoporosis [29,30]. Most publications report prevalence data for women only [31], and Table 3 Detailed NNS for the indication for osteoporotic treatment with regard to trauma energy or age groups (N= 478)

Age group

Gender No indication for treatment (DXA and FRAX)

Indication for t reatment (DXA and FRAX)

NNS 95% CI Trauma energy No indication for treatment (DXA

and FRAX)

Indication for treatment (DXA and FRAX)

NNS 95% CI

50+ Male 38 (58.5%) 27 (41.5%) 2.41 1.86–3.38 Low 21 (28%) 54 (72%) 1.39 1.22–1.62

Female 20 (30.8%) 45 (69.2%) 1.44 1.24–1.72 High 37 (67.3%) 18 (32.7%) 3.06 2.22–4.92 Total 58 (44.6%) 72 (55.4%) 1.81 1.56–2.13 Total 58 (44.6%) 72 (55.4%) 1.81 1.56–2.13 60+ Male 26 (59.1%) 18 (40.9%) 2.44 1.80–3.79 Low 34 (34.3%) 65 (65.7%) 1.52 1.33–1.78 Female 30 (34.9%) 56 (65.1%) 1.54 1.33–1.82 High 22 (71%) 9 (29%) 3.44 2.22–7.66 Total 56 (43.1%) 74 (56.9%) 1.76 1.53–2.07 Total 56 (43.1%) 74 (56.9%) 1.76 1.53–2.07 70+ Male 23 (79.3%) 6 (20.7%) 4.83 2.82–16.82 Low 36 (39.6%) 55 (60.4%) 1.65 1.42–1.98 Female 28 (33.7%) 55 (66.3%) 1.51 1.31–1.78 High 15 (71.4%) 6 (28.6%) 3.50 2.09–10.81 Total 51 (45.5%) 61 (54.5%) 1.84 1.57–2.21 Total 51 (45.5%) 61 (54.5%) 1.84 1.57–2.21 80+ Male 8 (34.8%) 15 (65.2%) 1.53 1.18–2.19 Low 25 (26.6%) 69 (73.4%) 1.36 1.21–1.55 Female 21 (25.3%) 62 (74.7%) 1.34 1.19–1.53 High 4 (33.3%) 8 (66.7%) 1.50 1.07–2.50 Total 29 (27.4%) 77 (72.6%) 1.38 1.23–1.56 Total 29 (27.4%) 77 (72.6%) 1.38 1.23–1.56

Total 194 (40.6%) 284 (59.4%) 1.68 1.57–1.82 194 (40.6%) 284 (59.4%) 1.68 1.57–1.82

NNSnumber needed to screen,DXAosteodensitometry,FRAXfracture risk assessment tool,DXA and FRAXcombined assessment of both DXA and FRAX

Table 4 Multivariate analysis (stepwise logistic regression) regarding the indication for the treatment of osteoporosis (DXA and FRAX;N= 478)

Variables B Wald p Odds ratio

exp (B)

99% CI General Improvement

Lower Upper Chi² p NagelkerkeR² Chi² p NagelkerkeR² Low energy 1.05 18.2 <0.001 2.84 1.76 4.59 40. 6 <0.001 0.11

Fracture femur 1.53 24.1 <0.001 4.63 2.51 8.53 60.1 <0.001 0.16 19.5 <0.001 0.05

Gender 0.92 16.7 <0.001 2.51 1.61 3.91 79.9 <0.001 0.21 19.8 <0.001 0.05

Fracture pelvis 1.37 6.7 0.010 3.92 1.39 11.04 86.0 <0.001 0.22 6.2 0.013 0.02

Fracture radius 0.56 4.6 0.032 1.75 1.05 2.93 90.7 <0.001 0.23 4.7 0.030 0.01

Constant −2.34 34.7 0.000 0.10

(7)

the same applies to FRAX threshold values [14,32]. The risk of fracture in men over 50 years of age has been described to range between 20 and 27%, which is lower than the comparable risk in women (50%) [33]. For Danish men >50 years, the prevalence of osteoporosis was estimated to be 17.7%

[34], but only 1.3% of men >60 years of age actually underwent the relevant pharmacotherapy. A recent report on over 129,000 US patients with a fragility fracture stated that only 10% of men underwent DXA testing and only 9.6%

received antiosteoporosis treatment [35]. A retrospective single-center evaluation on patients with distal radius fractures indicated that a male with a fragility fracture is 9.7 times less likely to undergo DXA testing than is a woman with the same fracture [36]. Our prospective findings favor screening for osteoporosis following non-vertebral fractures in both men and women.

Fifth, although we found a significant association in the rate of osteoporosis with increasing age in correlation with the literature, the NNS to detect an osteoporotic patient in the decennium 50–60 was only marginally worse than in higher decennia (e.g., 2.4 in men 50+ vs. 1.5 in men 80+).

Originally, this decision emanated from the assumption that a simple concept such as includingBall fracture patients aged 50 or older,^independently of trauma and gender, would be more feasible within the daily clinical routine of a teaching hospital.

The literature reports continuation of mostly very low screening rates for osteoporosis in fracture patients [10,35,37,38], partly due to the additional effort and specific knowledge required of the physicians involved [39–41]. No unique and well-accepted screening approach can be found in the current literature. The detailed criteria on when to screen which patients with which kind of examination vary widely, e.g., beginning at what age, for females depending on menopause, including what type of trauma or depending on which definition for fragility fracture, using different approaches to mea- sure BMD and additional risk scoring instruments such as FRAX, etc. [34,42–46]. Due to the limited number of patients in single decennia groups, especially if additionally stratified for gender or trauma energy, further interpretation is limited due to the possibility of random statistical findings. The example of a lower NNS found in HEF men aged 50+ compared to those aged 60+ or 70+ that is almost comparable to those aged 80+ underlines this statistical limitation. Overall, we are of the opinion that our preliminary data provide strong evidence for the simple approach of screening all fracture patients for osteoporosis beginning at the age of 50.

Sixth, we were interested in the differences between the NNS obtained for the more restricted assessment based on pathologic DXA measurement only (Tscore≤−2.5) to indicate treatment compared with the assessment that included FRAX scoring. With the inclusion of FRAX, the overall treatment prescription rate increased by almost 20%, i.e., the required NNS to initiate any treatment for osteoporosis would

decrease from 2.5 for DXA only to 1.7 with the additional use of FRAX. Originally, the definition of osteoporosis relied on the WHO-basedTscore of BMD. Following this definition, only subjects with aTscore at or below−2.5 were considered to have osteoporosis [47]. Almost all modern guidelines for osteoporosis [15,31] include additional risk factors and scores such as the FRAX to indicate therapy, but some omit DXA measurement of patients [32,42,48]. In the Rotterdam epide- miologic study, only 44% of women and 21% of men aged 55 and older with a non-vertebral fracture had a Tscore lower than−2.5 [47]. Follow-up studies on the long-term effects of such a more aggressive approach to indication are required to objectify the approach presented here. Given our NNS data and reports in recent literature suggesting that treatment of osteoporosis in patients with fragility fractures can reduce the risk of subsequent fractures by up to 50% and mortality rates by up to 30% [43,49,50], we believe that the affected fracture patients will profit importantly.

In addition, even though patients underwent detailed laboratory examination and a questionnaire for well-known risk factors of osteoporosis, we could not find any auxiliary relationship to the results of osteodensitometry or subsequent osteoporotic treatment. For example, neither 25-hydroxy vitamin D, TSH, PTH, nor self-declared alcohol intake showed a significant correlation to the obtained Tscore values. A comparison of these findings with the literature is difficult as most of the studies so far were undertaken in patients with well-known risk factors for osteoporosis and not in unselected fracture patients as in our study [51–54]. Kanis et al. address the problem of low sensitivity and low positive predictive values when screening a population of 50 years of age without a prevalent fracture [55].

Limitations

Until confirmed in other centers and larger cohorts, the findings from this preliminary study are limited to the setting of this monocentric evaluation and the patients under investigation. The public hospital is open to all patients, but there are no data available to confirm that the cohort under investigation is representative of the relevant population. The prospective study was performed with unselected fracture patients hospitalized at the trauma unit of a Swiss teaching hospital. Even though, hardly more than every second consecutive patient (56.3%) could be included. Although excluded patients were found to be older in univariate analysis, multivariable analysis revealed that it was mainly the fact that a person lived in a nursing home (and not their own home) that was associated with non-compliance with our standard procedure.

Additionally, conservative treatment of the indicator fracture correlated significantly with exclusion from the study. Given these comparative data and arguments based on the well- founded assumption that the NNS for the treatment of

(8)

osteoporosis in the group of more elderly and dependent patients would, in fact, be even higher than in our study cohort, we see no reason why this inclusion limitation should princi- pally detract from our major findings. Indeed, given the NNS for single risk groups, they should be considered even more impressive. Recent reports and population-based evaluations from different countries [35,56–58] show that only in about 10–30% of cases where treatment is indicated are the relevant diagnostics and/or treatments for osteoporosis actually imple- mented following fragility fractures or the diagnosis of osteoporosis. A recent review of randomized controlled trials investigating the professional care of patients with a fragility fracture and a variety of interventions to improve the implemen- tation of osteoporotic therapy revealed an overall 36% absolute increase in scanning rates for osteoporosis following ap- propriate interventions [59]. The diagnostic approach followed Swiss guidelines at the time of the study, which may differ in certain details from other countries. However, the major findings presented here should be largely independent of finer details in the diagnosis of osteoporosis but do need to be verified in larger studies. Especially subgroup findings such as the unexpectedly low NNS for male HEF patients aged 50 and older need to be reappraised because of the resulting small number of patients. Although we worked with a standardized stratification of trauma energy undertaken by one of the inves- tigators, we had to depend on the assumed best anamnestic information given by patients and/or any observers of trauma.

In summary, against the background of the recent literature, our preliminary findings reveal strong arguments in favor of performing standard screening for osteoporosis in all fracture patients aged 50 or older (women and men) independently of the trauma energy involved. The approach described here and the results obtained should be confirmed by larger studies and in other countries.

Acknowledgements The authors would like to thank all hospital col- laborators and Ms. J. Buchanan for editorial assistance.

Compliance with ethical standards The study was approved by the Cantonal Ethics Board and supported financially by the Research Fund of the Hospital.

Conflicts of interest None.

References

1. Jeray K (2015) Osteoporosis and fragility fractures. What progress have we made? J Bone Joint Surg Am 97:1553–1554

2. Neuerburg C, Stumpf U, Schmidmaier R, Kammerlander C, Pfeilschifter J, Mutscher W, Böcker W (2015) New DVO guideline for ostoeporosis management 2014 and its importance for trauma surgeons. Unfallchirurg 118:905–912

3. Bouillon R, Burckhardt P, Christiansen C et al (1991) Consensus development conference: prophylaxis and treatment of osteoporosis. Am J Med 90:107–110

4. Lippuner K (2009) Epidemiology of osteoporotic fractures in Switzerland. Revue Médicale Suisse 5:1304–1308

5. Svedbom A, Ivergard M, Hernlund E, Rizzoli R, Kanis J (2014) Epidemiology and economic burden of osteoporosis in Switzerland. Arch Osteoporos 9:186–187 (1–8)

6. Burge R, Dawson-Hughes B, Solomon D, Wong J, King A, Tosteson A (2007) Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Miner Res 22:465–475

7. Ahmed L, Center J, Bjoernerem A, Bluic D, Joakimsen R, Joergensen L, Meyer H, Nguyen N, Nguyen T, Omsland T, Stoermer J, Tell G, van Geel T, Eisman J, Emaus N (2013) Progressively increasing fracture risk with advancing age after ini- tial incident fragility fracture: the Tromsoe study. J Bone Miner Res 28:2214–2221

8. Mackey D, Lui L-Y, Cawthon P, Bauer D, Nevitt M, Cauley JA, Hillier T, Lewis C, Barrett-Connor E, Cummings S (2007) High trauma fractures and low bone mineral density in older women and men. JAMA 298:2381–2388

9. Sanders K, Pasco J, Ugoni A, Nicholson G, Seeman E, Martin T, Skoric B, Panahi S, Kotowicz M (1998) The exclusion of high trauma fractures may undererstimate the prevalence of bone fragility fractures in the community: the Geelon osteoporosis study. J Bone Miner Res 13:1337–1342

10. Dirschl D (2014) Shame on us! Men need osteoporsis care, too! J Bone Joint Surg Am 96:e186

11. Nguyen T, Sambrook P, Kelly P, Jones G, Lord S, Freund J, Eisman J (1993) Prediction of osteoporotic fractures by postural instability and bone density. BMJ 307:1111–1115

12. Lippuner K, von Overbeck J, Perretta L, Bosshard H, Jaeger P (1997) Incidence and direct medical costs of hospitalizations due to osteoporotic fractures in Switzerland. Osteoporos Int 7:414–425 13. DVO (2014) [Prophylaxis, diagnostics and therapy of osteoporosis].www.dv-osteologie.org/dvo_leitlinien/osteoporose-leitlinie- 2014

14. SVGO (2015) [Osteoporosis recommendations 2015—diagnostics, prevention, therapy].www.svgo.ch

15. Cosman F, de Beur S, Lewiecki E, Tanner B, Randall S, Lindsay R (2014) Clinician’s guide to prevention and treatment of osteoporosis. Osteoporos Int 25:2359–2381

16. Kanis J, the WHO study group (1994) Assessment of fracture risk and its application to screening for postmenopausal osteoporosis:

synopsis of a WHO report. Osteoporos Int 4:368–381

17. Charlson M, Charlson R, Peterson J, Marinopoulos S, Briggs W, Hollenberg J (2008) The Charlson comorbidity index is adapted to predict costs of chronic disease in primary care patients. J Clin Epidemiol 61:1234–1240

18. Charlson M, Szatrowski T, Peterson J, Gold J (1994) Validation of a combined comorbidity index. J Clin Epidemiol 47:1245–1251 19. Warriner A, Patkar N, Yun H, Delzell E (2011) Minor, major, low-

trauma, and high-trauma fractures: what are the subsequent fracture risks and how do they vary? Curr Osteoporos Rep 9:122–128 20. Huisa B, Stemer A, Zivin J (2010) Atorvastatin in stroke: a review

of SPARCL and subgroup analysis. Vasc Health Risk Manag 6:

229–236

21. Hernlund E, Svedbom A, Ivergard M, Compston J, Cooper C, Stenmark J, McCloskey E, Jönsson B, Kanis J (2013) Osteoporosis in the European Union: medical management, epidemiology and economic burden: a report prepared in collaboration with the International Osteoporosis Foundation (IOF) and the European Federation of Pharmaceutical Industry Associations (EFPIA). Arch Osteoporos 8:136

(9)

22. National Osteoporosis Foundation (2014) Clinician’s guide to prevention and treatment of osteoporosis. National Osteoporosis Foundation 1:1–55

23. Lofthus C, Osnes E, Meyer H, Kristiansen I, Nordsletten L, Falch J (2006) Young patients with hip fracture: a population-based study of bone mass and risk factors for osteoporosis. Osteoporos Int 17:

1666–1672

24. Hedlund R, Lindgren U (1987) Trauma type, age, and gender as determinants of hip fracture. J Orthop Res 5:242–246

25. Robinovitch S, Hayes W, McMahon T (1991) Prediction of femoral impact forces in falls on the hip. J Biomech Eng 113:366–374 26. Chiu J, Robinovitch S (1998) Predicition of upper extremitiy im-

pact forces during falls on the outstretched hand. J Biomech 31:

1169–1176

27. Choi W, Cripton P, Robinovitch S (2015) Effects of hip abductor muscle forces and knee boundary conditions on femoral neck stresses during simulated falls. Osteoporos Int 26:291–301 28. Mackey D, Black D, Bauer D, McCloskey E, Eastell R, Mesenbrink

P, Thompson J, Cummings S (2011) Effects of antiresorptive treatment on nonvertebral fracture outcomes. J Bone Miner Res 26:

2411–2418

29. Szulc P, Munoz F, Duboeuf F, Marchand F, Delmas P (2005) Bone mineral density predicts osteoporotic fractures in elderly men: the MINOS study. Osteoporos Int 16:1184–1192

30. Mussolino M, Looker A, Madans J, Langlois J, Orwoll E (1998) Risk factors for fip fracture in white men: the NHANES I epidemi- ologic follow-up study. J Bone Miner Res 13:918–924

31. Kanis J, McCloskey E, Johansson H, Cooper C, Rizzoli R, Reginster J-Y, ESCEO, IOF (2012) European guidance for diagnosis and management of osteoporosis in postmenopausal women.

Osteoporos Int 24(1):23–57

32. Kanis J, Harvey N, Cooper C, Johansson H, Odén A, McCloskey E (2016) A systematic review of intervention thresholds based on FRAX. Arch Osteoporos 11:25

33. Frost M, Gudex C, Rubin K, Brixen K, Abrahamsen B (2012) Pattern of the use of DXA scans in men: a cross sectional population-based study. Osteoporos Int 23:183–191

34. Vestergaard P, Rejnmark L, Mosekilde L (2005) Osteoporosis is markedly underdiagnosed: a nationwide study from Denmark.

Osteoporos Int 16:134–141

35. Balasubramanian A, Tosi L, Lane J, Dirschl D, Ho P, O’Malley C (2014) Declining rates of osteoporosis management following fragility fractures in the U.S., 2000 through 2009. J Bone Joint Surg Am 96:e52

36. Harper C, Fitzpatrick S, Zurakowski D, Rozental T (2014) Distal radial fractures in older men. J Bone Joint Surg Am 96:1820–1827 37. Giangregorio L, Papaioannou A, Cranney A, Zytaruk N, Adachi J (2006) Fragility fractures and the osteoporosis care gap: an international phenomenon. Semin Arthritis Rheum 35:293–305 38. Leslie W, Giangregorio L, Yogendran M, Azimaee M, Morin S,

Metge C, Caetano P, Lix L (2012) A population-based analysis of the post-fracture care gap 1996-2008: the situation is not improving.

Osteoporos Int 23:1623–1629

39. Edwards B, Koval K, Bunta A, Genuario K, Hahr A, Andruszyn L, Williams M (2011) Adressing secondary prevention of osteoporosis in fracture care: follow-up toBown the bone^. J Bone Joint Surg Am 93:e87(1)–e87(7)

40. Chenot R, Scheidt-Nave C, Gabler S, Kochen M, Himmel W (2007) German primary care doctor’s awareness of osteoporosis and knowledge of national guidelines. Exp Clin Endocrinol Diabetes 115:584–589

41. Rosenwasser M, Cuellar D (2016) Medical management of osteoporosis and the surgeons’role. Injury 47(Supplement 1):S62–S64

42. Kanis J, Harvey N, Johansson H, Odén A, Leslie W, McCloskey E (2015) FRAX and fracture prediction without bone mineral density.

Climacteric 18:2–9

43. Hochberg M (2000) Preventing fractures in postmenopausal women with osteoporosis. Drugs Aging 17:317–330

44. Papaioannou A, Morin S, Cheung A, Aktinson S, Brown J, Feldman S, Hanley D, Hodsman A, Jamal S, Kaiser S, Kvern B, Siminoski K, Leslie W (2010) 2010 clinical practice guidelines for the diagnosis and management of osteoporosis in Canada: summary. CMAJ 182(17):1864–1873

45. Leslie W, Schousboe J (2011) A review of osteoporosis diagnosis and treatment options in new and recently updated guidelines on case finding around the world. Curr Osteoporos Rep 9:129–140 46. Kanis J, Compston J, Cooper C, Harvey N, Johansson H, Odén A,

McCloskey E (2016) SIGN guidelines for Scotland: BMD versus FRAX versus QFracture. Calcif Tissue Int 98:417–425

47. Schuit S, van der Klift M, Weel A, de Laet C, Burger H, Seeman E, Hofman A, Uitterlinden A, van Leeuwen J, Pols H (2004) Fracture incidence and association with bone mineral density in elderly men and women: the Rotterdam study. Bone 34:195–202

48. Kanis J, McCloskey E, Branco J, Brandi M, Dennison E, Devogelaer J, Ferrari S, Kaufman J, Papapoulos S, Reginster J, Rizzoli R (2014) Goal-directed treatment of osteoporosis in Europe. Osteoporos Int 25:2533–2543

49. Grey A, Bolland M (2013) The effect of treatments for osteoporosis on mortality. Osteoporos Int 24:1–6

50. Bolland M, Grey A, Gamble G, Reid I (2010) Effect of osteoporosis treatment on mortality: a meta-analysis. The Journal of Clinical Endocrinology & Metabolism 95:1174–1181

51. Hudec S, Camacho P (2012) Secondary causes of osteoporosis.

Endocr Pract 19:120–128

52. Waring A, Harrison S, Fink H, Samuels M, Cawthon P, Zmuda J, Orwoll E, Bauer D (2013) A prospective study of thyroid function, bone loss, and fractures in older men: the MrOS study. J Bone Miner Res 28:472–479

53. Kushnir-Sukhov N, Brittain E, Reynolds J, Akin C, Metcalfe D (2006) Elevated tryptase levels are associated with greater bone density in a cohort of patients with mastocytosis. Int Arch Allergy Immunol 139:265–270

54. Bours S, van Geel T, Geusens P, Janssen M, Janzing H, Hoffland GA, Willems P, van den Bergh J (2011) Contributors to secondary osteoporosis and metabolic bone disease in patients presenting with clinical fracture. The Journal of Clinical Endocrinology &

Metabolism 96:1360–1367

55. Kanis, JA and on behalf of the WHO Scientific Group (2007) Assessment of osteoporosis at the primary health care level.

www.shef.ac.uk/FRAX/pdfs/WHO_Technical_Report.pdf 56. Vogel T, Kampmann P, Bürklein D, Böhm H, Ockert B, Kirchhoff

C, Kanz K, Pfeifer K, Mutschler W (2008) Reality of treatment of osteoporotic fractures in German trauma department. Unfallchirurg 111:869–877

57. Suhm N, Lamy O, Lippuner K (2008) Management of fragility fractures in Switzerland: results of a nationwide survey. Swiss Med Wkly 138:674–683

58. McLellan A, Wolowacz S, Zimovetz E, Beard S, Lock S, McCrink L, Adekunle F, Roberts D (2011) Fracture liaison services for the evaluation and management of patients with osteoporotic fracture:

cost-effectiveness evaluation based on data collected over 8 years of service provision. Osteoporos Int 22:2083–2098

59. Little E, Eccles M (2010) A systematic review of the effectivness of interventions to improve post-fracture investigation and managment of patients at risk of osteoporosis. Implement Sci 5:

80–97