hta dexa intro€¦ · This report should be cited as: Amicosante AMV, Bernardini F, Cavallo A,...

HTA REPORT

Technologies for the

identification of osteoporosis

This report should be cited as: Amicosante AMV, Bernardini F, Cavallo A, Cerbo M, Jefferson T,Lo Scalzo A, Ratti M. Agenas HTA Report – Technologies for the identification of osteoporosis.Rome, July 2009.

Contributions

Authors:Anna Maria Vincenza Amicosante, Fabio Bernardini, Antonella Cavallo, Marina Cerbo, TomJefferson, Alessandra Lo Scalzo, Marco Ratti.

Agenas, Agenzia nazionale per i servizi sanitari regionali Sezione Iss (Innovazione, sperimentazione e sviluppo), Rome (Italy)

Corresponding author:Anna Maria Vincenza Amicosante ([email protected])

Experts:Carlina Albanese

MD Senior Consultant (Chief of the study group of osteoporosis and bone pathology)Department of radiologic science - Policlinico Umberto I, Rome (Italy)

Antonio Migliore

Consultant (Biomedical Engineer)Agenas, Agenzia nazionale per i servizi sanitari regionali, Rome (Italy)

External Reviewers:Maryann Napoli

Center for Medical ConsumersNew York,U.S.A.

Bibliographic searchesFabio Bernardini

Agenas, Agenzia nazionale per i servizi sanitari regionali

Editing and layoutDario Fella

Agenas, Agenzia nazionale per i servizi sanitari regionali

Acknowledgements:The Authors and Agenas would like to thank all the responding centres (see Appendix 8) who

contributed to this HTA report by participating to the survey and all the responding regions (seepar. 6.1) who contributed to this HTA report by sending the administrative data of outpatient pro-cedures for densitometric technologies.

V

HTA REPORT

Technologies for the

identification of osteoporosis

Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .VIII

Prefazione . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .IX

Executive summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .XI

Sintesi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .XIII

1. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11.1 Clinical problem and target population . . . . . . . . . . . . . . . . . . . . . . . .11.1.1 Clinical problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11.1.2 Target population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

2. Technology, procedure, measurements and alternatives . . . . . . . . . . . .32.1 DEXA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32.2 The DEXA procedure and measurements . . . . . . . . . . . . . . . . . . . . .32.3 The alternative technologies

and the choice of QUS as comparator . . . . . . . . . . . . . . . . . . . . . . . .42.4 DEXA and QUS:

different measurements by different technologies . . . . . . . . . . .5

VI

3. The marketing status of DEXA and QUSand current reimbursement arrangements . . . . . . . . . . . . . . . . . . . . . . . . . .73.1 DEXA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73.2 QUS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7

4. Report’s objectives: policy and research question . . . . . . . . . . . . . . . . . .9

5.Systematic review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115.1 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

5.2.1 Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115.2.2 Accuracy of DEXA versus QUS

in the identification of osteoporosis . . . . . . . . . . . . . . . . . . . . .125.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

6. Context specific analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .156.1 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .156.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18

6.2.1 Analysis of administrative data . . . . . . . . . . . . . . . . . . . . . . . . . .186.2.2 Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25

6.2.2.1 Characteristics of procedures andvolumes of activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25

6.2.2.2 Characteristics of procedures for postmenopausal women . . . . . . . . . . . . . . . . . . . . .27

6.2.2.3 Organisational and economic aspects of technologies and procedures . . . . . . . . . . . . . . . . . . .28

7. Cost effectiveness analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .337.1 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .337.2 Efficacy of DEXA and QUS in the identification

of osteporosis in the target population . . . . . . . . . . . . . . . . . . . . . .337.3 Costs analysis for DEXA and QUS procedures . . . . . . . . . . . . . .347.4 Cost effectiveness analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .367.5 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38

8. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39

9. Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41

10. Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43

11. Competing interests declaration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47

Appendices1 - Technologies for bone measurements . . . . . . . . . . . . . . . . . . . . . . . . . .512 - Search Strategy by database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .573 - QUADAS tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .594 - List of excluded studies with reasons for exclusion . . . . . . . . . . .615 - Tables of evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .776 - QUADAS for included studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .817 - Survey questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .898 - Responding centres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101

Acronyms and abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103

VII

Foreword

This year Agenas has produced two Health Technology Assessment reports for the MinisterialCommittee on Medical Devices (CUD). The present report analyses the best and most appropria-te evidence of costs and effects of what has become a commonly used test in the Italian NHS andworldwide to identify and assess osteoporosis: dual-energy x-ray absorptiometry (DEXA) and itsdirect competitor dual Quantitative Ultra Sound (QUS).

The report and its content are as always the product of a long and laborious process involvingconsultation with experts, referees, producers and other stakeholders.

The report is based on a direct study question of which of the two technologies performs bestto identify osteoporosis in women aged 50 or over. One of the unique features of the report is theattempt at finding out the volume of activity and direct costs of the two procedures from a ran-dom sample of centres using one or the other or both of them. This allowed Agenas to constructa cost effectiveness ratio which confirms that DEXA is the dominant technology. It is Interestingto note that the current release of DEXA seems to be appropriate. One of our referees, MaryannNapoli who is associate director of the Center for Medical Consumers in NewYork City, pointed outthat osteoporosis is an age related risk factor, not a disease or a pre-morbid condition. Her fear isthat report may be seen as an encouragement towards the ethically ambiguous introduction ofscreening healthy women (and subsequent drug intervention in a proportion of them), when infact they are not either symptomatic nor diseased.

It is for this reason that the term diagnosis, which is usually associated with disease has beeneliminated from the report.

With this premise I wish all readers an interesting time in reading the report.

Fulvio Moirano

Executive Director age.na.s

VIII

IX

Prefazione

L’Agenas quest’anno ha prodotto due report di Health Technology Assessment per laCommissione ministeriale sui dispositivi medici (CUD). Questo report analizza le migliori e piùappropriate evidenze dei costi e degli effetti del test che è diventato di uso comune nel Serviziosanitario italiano e in tutto il mondo per individuare e valutare l'osteoporosi: l’assorbimetria a raggiX a doppia energia (DEXA) e il suo diretto comparatore l’ultrasonografia quantitativa (QUS).

Il report e il suo contenuto sono come sempre il prodotto di un lungo e laborioso processo diconsultazione con esperti, referee, produttori e altri stakeholders.

Il report si basa su uno studio diretto a capire quale delle due procedure tecnologiche sia lamigliore per individuare l'osteoporosi nelle donne di età uguale o superiore ai 50 anni. Una dellecaratteristiche del report è la rilevazione del volume delle attività e dei costi diretti delle due pro-cedure, attraverso un campione ragionato di centri che utilizzano una o l'altra o entrambe. Ciò haconsentito all’Agenas di costruire un rapporto costo-efficacia che conferma che la DEXA è la tec-nologia dominante. È interessante, inoltre, sottolineare che l’attuale erogazione della DEXA, sem-bra avvenire appropriatamente. Uno dei nostri referee, Maryann Napoli, che è il Direttore associa-to del Center for Medical Consumers di New York City, ha sottolineato che l'osteoporosi è un fat-tore di rischio connesso all’età, non è una malattia o una condizione premorbosa. Il suo timore èche il report possa essere considerato come un incoraggiamento verso l’introduzione, eticamentedubbia, dello screening su donne sane (e il conseguente intervento farmacologico per una partedi queste), mentre in realtà non sono né sintomatiche né malate.

È per questo motivo che il termine diagnosi, che è di solito associato alla malattia, è stato eli-minato dal report.

Con questa premessa auguro a tutti i lettori una interessante lettura.

Fulvio Moirano

Direttore age.na.s

XI

Executive summary

One-liner We assessed accuracy and cost effectiveness of dual-energy x-ray absorptiometry (DEXA)

compared with other procedures to identify osteoporosis in the Italian postmenopausal femalepopulation.

Background Osteoporosis is widespread in postmenopausal women and causes a reduction both in the

quality and density of the bone increasing the risk of fracture. The most common fractures asso-ciated with osteoporosis occur in the wrist, hip and spine and their consequences are usually aloss of independence, social isolation, a deterioration of the quality of life, surgery and in somecases death. DEXA is considered the gold standard in assessing the bone mineral density. Othertechnologies used are Quantitative Computed Tomography (QCT), Quantitative Ultra Sound(QUS), Radiographic Absorptiometry (RA).

Objectives The objectives of this report are: to identify the most widely used technologies for the identi-

fication and assessment of osteoporosis in the target population in our country; to accuratelydescribe the acquisition and use of the technology and its alternative in the Italian context; toassess the evidence for the chosen technologies, to perform a cost-effectiveness analysis compa-ring DEXA with the most widely used alternatives used in the target population.

Methods We analysed available Italian data to gather information on the diffusion, volumes, and rele-

vant costs of the various technologies in Italy, and performed a systematic review of their accura-cy in the identification of osteporosis and a cost effectiveness analysis.

Results DEXA is the most used device at outpatient level for the identification of osteoporosis in our

target population and has a higher imaging efficacy when compared with QUS (its most likely com-petitor in 14 Italian regions). We identified 7 studies comparing DEXA with QUS in postmenopau-sal women. We did not carry out a meta-analysis. When using basic DEXA and QUS equipment,our economic evaluation based on over 12 and a half million women shows that DEXA is the domi-nant procedure for identifying osteoporosis in a general population of postmenopusal women(incremental cost effectiveness ratio €1.095).

XII

ConclusionDEXA is the dominant technology. Should improvements arise, this report should be updated.

XIII

Sintesi

Problema clinico e popolazione targetL’osteoporosi è una condizione caratterizzata da una diminuzione della massa ossea dello sche-

letro. Le ossa diventano più porose e fragili e sono a più alto rischio di frattura. Le fratture piùcomunemente associate all’osteoporosi riguardano il polso, l’anca e la colonna vertebrale possonocausare una perdita di autonomia, peggioramento della qualità della vita, isolamento sociale, inter-venti chirurgici e, in alcuni casi, la morte.

Si classificano due tipologie di osteoporosi:

• l’osteoporosi primaria causata dall’invecchiamento, dalla menopausa e da alcuni fat-tori legati allo stile di vita come il fumo, l’alcol, la dieta e l’inattività fisica;

• l’osteoporosi secondaria causata da alcune malattie e/o dall’uso di alcuni farmaci1.

L’osteoporosi primaria può interessare sia la popolazione femminile che quella maschile, ma acausa delle modificazioni ormonali indotte dalla menopausa, essa appare più diffusa tra la popo-lazione femminile in postmenopausa (Cooper et al. 1992). Dato questo maggior rischio per ledonne in postmenopausa, il presente report ha individuato questo sottogruppo come propriapopolazione target, al fine di valutare le tecnologie più idonee per una accurata identificazione diquesta condizione.

Descrizione della tecnologia e della procedura Secondo l’OMS (1994) il gold standard per l’identificazione dell’osteoporosi è la densitometria

ossea con tecnica di assorbimento a raggi X (DEXA). Essa viene eseguita in ambiente dedicato,con l’ausilio di un tecnico radiologo e non è trasportabile. Tuttavia esistono altre tecnologie chepossono essere utilizzate con questo scopo, quali la densitometria ossea con tac (QCT), la densi-tometria ossea ad ultrasuoni (QUS), la radiografia (RA), la risonanza magnetica (MRI), la tecnicadi assorbimento a fotone singolo o doppio (SFA o DPA) .

Nella scelta del comparatore rispetto al quale valutare l’accuratezza di identificazione e la costoefficacia della DEXA, abbiamo tenuto conto sia di aspetti specifici al nostro contesto italiano (datidi utilizzo regioni italiane), sia della letteratura di settore. Data la maggiore propensione delleRegioni all’utilizzo della QUS, dopo la DEXA, e tenuto conto che in letteratura si enfatizzano la mag-giore trasportabilità, la assenza di problematiche relative alla sicurezza, e la facilità nell’esecuzio-ne della QUS (MacLaughlin, 2006), questa è stata la tecnologia scelta per l’analisi comparativa.

Obiettivi del reportLa nostra Policy Question (PQ) è volta a capire se la DEXA sia la tecnica più costo efficace per

l'identificazione dell’osteoporosi nelle donne in postmenopausa, se confrontata con le altre alter-native più comunemente utilizzate in Italia.

XIV

Per rispondere a questa PQ abbiamo identificato le seguenti domande di ricerca:

• Qual’è l'evidenza scientifica disponibile circa l’accuratezza e la sicurezza della DEXAvs le alternative più diffuse nel contesto italiano, per l’identificazione dell’osteoporosiprimaria nella popolazione target?

• Quali sono le tecnologie più diffuse per l'identificazione dell’osteoporosi in Italia,come vengono utilizzate e quali sono i loro costi?

• Qual è la tecnologia più costo-efficace per l’identificazione dell’osteoporosi primarianel nostro target?

Il report ha quindi quattro principali obiettivi:

1) capire quali sono le tecnologie più diffuse in Italia per l'identificazione dell‘osteoporosinella popolazione target;

2) individuare, acquisire, aggiornare e valutare le evidenze sulle tecnologie selezionate,prendendo in considerazione l’accuratezza nell’identificazione dell’osteoporosi lasicurezza e l’impatto economico;

3) descrivere l'acquisizione e l'uso della tecnologia e le sue alternative nel contesto italiano;

4) eseguire una analisi costo-efficacia confrontando il gold standard, DEXA, con l’alter-nativa più diffusa.

MetodiL’accuratezza della DEXA confrontata con quella della QUS è stata valutata attraverso una revi-

sione sistematica in cui abbiamo incluso tutti gli studi primari e di sintesi, che comparassero DEXAcon QUS in donne sane, caucasiche in postmenopausa, nel periodo di riferimento 1998-2008. Lavalutazione di qualità degli studi inclusi è stata effettuata tramite l’applicazione dello strumentoQuality assessement of diagnostic accuracy studies (QUADAS).

Sono stati poi analizzati i dati amministrativi delle Regioni italiane per capire quali fossero letecnologie più diffuse per l'identificazione dell’osteoporosi primaria nella popolazione oggetto distudio e le loro caratteristiche specifiche. Le potenziali informazioni presenti nel database SDO,non sono invece state utilizzate, poiché l'ICD9CM riportano dati aggregati per tutte le tecnichedensitometriche (ICD9-CM: 88.79 - altra diagnostica ad ultrasuoni e 88.98 - densitometria ossea).

Poiché il sistema informativo sanitario italiano non ha un database consolidato sulle prestazio-ni ambulatoriali abbiamo coinvolto le Regioni, chiedendo loro l’invio dei dati sulle prestazioni densi-tometriche, per gli anni 2006 e 2007, relative ai codici 88.99.1 (densitometria ossea con tecnica diassorbimento a fotone singolo o doppio - polso o caviglia); 88.99.2 (densitometria ossea con tec-nica di assorbimento a raggi x - lombare, femorale, ultradistale) 88.99.3 (densitometria ossea contecnica di assorbimento a raggi x - total body); 88.99.4 (densitometria ossea con tc – lombare);88.99.5 (densitometria ossea ad ultrasuoni) (Decreto ministeriale - 22 Luglio 1996). Per avere piùinformazioni su caratteristiche, modalità di utilizzo e costi di DEXA e QUS nei vari centri italiani incui vengono utilizzate queste metodiche, è stata condotta una indagine conoscitiva. Sulla base deidati amministrativi, ad un campione ragionato di centri (36 unità) che eseguono solo DEXA, sia QUSche DEXA e solo QUS, è stato inviato un questionario per rilevare, relativamente all’anno 2008:

• le caratteristiche delle procedure (totale dei volumi, numero di pazienti, numero diesami per osteoporosi primaria, tipo di esame, volumi suddivisi per età);

XV

• le caratteristiche delle procedure su donne in postmenopausa (numero totale delledonne e delle donne in postmenopausa, ragioni, risultati);

• gli aspetti economici, organizzativi e delle procedure (tempo dedicato dal personalesanitario, quantità dei materiali di consumo, tipo e prezzo dei dispositivi ecc.).

Utilizzando le informazioni dell’indagine conoscitiva su costi e modalità di esecuzione delle pro-cedure, e integrandoli con le informazioni di accuratezza ottenute dallo studio di Clowes, abbiamoeseguito un’analisi costo-efficacia volta a confrontare, nella nostra popolazione target, il gold stan-dard DEXA versus QUS. Il punto di vista adottato è il punto di vista del SSN. L’analisi di sensibilitàè stata eseguita considerando una variazione dei costi minimi e massimi, come da indagine cono-scitiva, e della prevalenza di osteoporosi primaria nella popolazione target (9,8% desunta dallo stu-dio di Clowes, 2006 e 20% desunta dai dati di prevalenza riportati dallo studio italiano ESOPO,2004).

RisultatiNella nostra revisione sistematica, da un totale di 184 studi identificati dalla strategia di ricer-

ca, abbiamo incluso 7 studi che comparavano DEXA e QUS nelle donne in postmenopausa. Gli studiindividuati sono tutti di tipo osservazionale e la qualità e l’eterogeneità degli studi inclusi non ha ciha permesso di effettuare metanalisi. Tranne gli studi di Scott 2004, Frost 2001 e Clowes 2006 glistudi rimanenti sono basati su gruppi di pazienti di numerosità modesta. Nello studio di Clowes,l’unico dei tre ad avere tutte le caratteristiche adatte al nostro obiettivo valutativo, la DEXA effet-tuata all’anca ha un’efficacia identificativa maggiore della QUS Achilles BUA (sensibilità 96%, spe-cificità 55%) e DBM Ad SOS (sensibilità 96%, specificità 33%). Relativamente alla sicurezza, neglistudi individuati non si riportano valutazioni specifiche relative a questa dimensione. D’altra partesi stima che l’esposizione ai raggi X durante la procedura DEXA non ponga particolari problemi peril paziente: Thomas e colleghi valutano infatti inferiore allo 0,0001% l’aumento del rischio di svilup-pare un cancro per ogni scansione vertebrale DEXA.

Nella ricostruzione e analisi dei dati amministrativi abbiamo avuto un feedback completo da 14Regioni e Province Autonome (P.A.) e sono risultati aggregabili solo i database relativi all’anno 2007.Dalla loro analisi è emerso che nelle Regioni si eseguono principalmente esami DEXA, QCT e QUS.Per la DEXA multi sito (lombare, femorale e ultradistale) le prestazioni risultano essere 885.434, perla DEXA total body sono 40.738, mentre per la QTC 31.057 e per la QUS 26.777. La QUS risultaperò, in termini relativi, la procedura che dopo la DEXA è più utilizzata dalla maggior parte delleRegioni (8/12). Focalizzando l’attenzione sull’utilizzo di DEXA e QUS nella nostra popolazione tar-get, i dati confermano il prevalente uso della DEXA multi sito, seguito dalla DEXA total body e QUS.I tassi grezzi regionali nell’età immediatamente successiva alla menopausa (minimo valore 0.1%delle Marche e massimo 14% Lazio) per la DEXA mostrano livelli differenziati di utilizzo regionaleche si annullano nelle classi delle ultraottantacinquenni. Per la QUS i tassi di utilizzo grezzi per etàsono bassi, con un moderato trend crescente fino alla classe di età 65-75 e un moderato trenddecrescente nelle classi successive. I tassi regionali standardizzati per età mostrano una rilevantevariabilità interregionale per ambedue le procedure, registrando i più alti tassi di utilizzo della DEXAnel Lazio, Liguria e Toscana, mentre per la QUS emerge la PA di Bolzano.

L’indagine conoscitiva condotta su 36 centri italiani ha avuto una percentuale di ritorno del47,2%: 17 centri hanno risposto ma solo 16 sono stati in grado di fornire i dati. Dei 16 centri rispon-

XVI

denti, 6 centri (37,5%) effettuavano solo DEXA; 4 centri (25%) effettuavano solo QUS; 6 centri(37,5%) effettuavano entrambe le procedure. Nell’anno 2008 le strutture censite hanno erogato93.862 procedure (tra DEXA e QUS). La maggior parte delle procedure DEXA (88%) è stata esegui-ta in ambito ambulatoriale così come le procedure QUS (92%). La principale motivazione per l’ese-cuzione di entrambe le procedure è stata l’identificazione dell’osteoporosi primaria (rispettivamente80% e 84% per DEXA e QUS).

Abbiamo stimato, attraverso i dati forniti dai centri rispondenti, il costo medio diretto di un sin-golo esame DEXA (tenendo in considerazione i costi del personale, del materiale di consumo e delleapparecchiature) pari a €30,81 (range: €14,16-€49,33) e il costo medio diretto di un singolo esameQUS pari a €26,43 (range: €16,04-€33,37).

L’analisi costo efficacia ha evidenziato infine che, quando viene utilizzato il modello base diDEXA e QUS il nostro modello economico, applicato ad una popolazione di circa 12 milioni didonne, mostra che la DEXA è la procedura dominante per l’identificazione dell’osteoporosi in unapopolazione di donne in post menopausa (ICER €1.095).

DiscussioneIl presente studio ha dimostrato che ad oggi la DEXA risulta essere la tecnologia dominante

per l’identificazione dell’osteoporosi in Italia.Anche se la QUS potrebbe essere intesa come un’al-ternativa più economica e sicura alla DEXA, le performance cliniche individuate sono inferiori.Sebbene questo non fosse tra gli obiettivi principali del presente report, intendiamo segnalare,inoltre, che la QUS non si mostra adatta a programmi di screening poiché caratterizzata da unabassa specificità che comporta alte percentuali di falsi positivi.

RaccomandazioniRaccomandiamo che non siano apportati cambiamenti al corrente sistema di erogazione della

DEXA. Suggeriamo che sia istituito un audit (rolling audit) presso i centri che erogano la DEXA perla valutazione del fabbisogno, equità ed appropriatezza. Nel caso di miglioramenti delle performan-ce della QUS o di future ulteriori alternative, il presente report dovrebbe essere aggiornato perconsiderare tali cambiamenti.

1

1. Background

1.1 Clinical problem and target population

1.1.1 Clinical problem

Osteoporosis is a condition that causes a reduction of the density and quality of bone. Thebone becomes more porous and fragile and has a higher risk of fracture. Two different types ofosteoporosis have been defined.

• Primary osteoporosis is caused by aging, menopause, and certain lifestyle factorssuch as smoking, alcohol, diet, and physical inactivity.

• Secondary osteoporosis is due to disease and/or use of drugs1.

Our focus is on primary osteoporosis, which is common in the natural aging process, althoughit does not affect everyone in the same manner. Osteoporosis is usually associated with a higherrisk of fracture and the most common fractures associated with it occur at the wrist, hip andspine. The incidence of the latter increases with age in both men and women, and can result inserious consequences such as loss of independence, social isolation, a deterioration of quality oflife, surgery and death. Mortality from hip fractures within a year appears to be 10/15 per centhigher than in a corresponding group of the same gender and age without hip fracture1,2. Anessential element for fracture risk assessment is bone densitometry3. The most widely used cri-teria for assessing osteoporosis are those promulgated by the World Health Organisation (WHO).WHO recommends that bone measurements are expressed as T scores4 (see paragraph 2.2 fordefinitions).

1.1.2 Target population

Due to the higher risk for postmenopausal women to develop primary osteoporosis5, we willassess the relevant technologies for the identification of osteoporosis in this target population. Inthis population the rate of bone loss accelerates because the level of oestrogens, produced by theovaries greatly declines after menopause. In Italy the relative prevalence of osteopenia and osteo-porosis in women is 42,3% and 21,7% and in man of the 34,3% and 18% respectively. In themale population these percentages are relatively stable from 60 to 79 years. In the female thereis a progressive increase from 40 to 79 years. This implies that 86% of the elderly female popu-lation can be defined as osteopenic or osteoporotic, but it seems that a large number of casesremain undiagnosed6. Prevalence data on postmenopausal women in Italy are not available. Weestimated the number of our target population from ISTAT data (the Italian central institute of sta-tistics), considering the number of women that are very unlikely to be in a fertile age: all womenaged 50 or more (see Table 1.1).

2

Table 1.1: Italian population broken down by region, sex and age (ISTAT 2007)

Region

Female Male

Percentage Absolutevalue Percentage Absolute

value

Aged lessthan 50

Aged 50 ormore Total Total Aged less

than 50Aged 50 or

more Total Total

PIEMONTE 54.9 45.1 100.0 2241506 60.5 39.5 100.0 2111322

VALLE D'AOSTA 57.7 42.3 100.0 63379 63.0 37.0 100.0 61433

LOMBARDIA 58.4 41.6 100.0 4885089 64.6 35.4 100.0 4660352

TRENTINO ALTO ADIGE 61.8 38.2 100.0 505361 67.1 32.9 100.0 489342

VENETO 59.1 40.9 100.0 2435497 64.9 35.1 100.0 2338057

FRIULI VENEZIA GIULIA 53.7 46.3 100.0 626058 60.5 39.5 100.0 586544

LIGURIA 50.2 49.8 100.0 844472 56.9 43.1 100.0 763406

EMILIA ROMAGNA 55.6 44.4 100.0 2168358 61.5 38.5 100.0 2054906

TOSCANA 54.5 45.5 100.0 1882121 60.3 39.7 100.0 1756090

UMBRIA 55.1 44.9 100.0 450762 60.5 39.5 100.0 422205

MARCHE 56.4 43.6 100.0 788032 61.8 38.2 100.0 748066

LAZIO 59.4 40.6 100.0 2853033 64.5 35.5 100.0 2640275

ABRUZZO 58.1 41.9 100.0 672751 62.9 37.1 100.0 637046

MOLISE 57.6 42.4 100.0 164152 62.5 37.5 100.0 155922

CAMPANIA 65.8 34.2 100.0 2977552 69.9 30.1 100.0 2812635

PUGLIA 62.5 37.5 100.0 2093744 66.8 33.2 100.0 1976125

BASILICATA 60.6 39.4 100.0 301176 64.9 35.1 100.0 290162

CALABRIA 62.4 37.6 100.0 1023372 66.0 34.0 100.0 974680

SICILIA 62.4 37.6 100.0 2591683 66.7 33.3 100.0 2425178

SARDEGNA 60.5 39.5 100.0 844748 64.9 35.1 100.0 814695

Italy 59.0 41.0 100.0 30412846 64.3 35.7 100.0 28718441

3

2. Technology, procedure, measurements and alternatives

2.1 DEXAThe dual-energy x-ray absorptiometry (DEXA or also known as DXA) was first developed

approximately 20 years ago7 and today is a well established technology to measure Bone mineraldensity (BMD) (see Appendix 1) and the gold standard for diagnosing osteoporosis4. DEXA can beconsidered as an evolution of previous technologies (e.g. single-photon absorptiometry, SPA, anddouble-photon absorptiometry DPA) that used the different energy attenuation of bone and softtissue to measure the BMD. While SPA and DPA, use isotope sources to obtain a single or doublephoton beam respectively, DEXA utilises an x-ray source to generate images and to perform mea-surements. The main advantage is a stable incident radiation (no decay, as with an isotope sour-ce). Moreover, all the risks and costs related to the management of the isotope source are avoi-ded (e.g. radioactive waste disposal).

2.2 The DEXA procedure and measurementsDuring a DEXA scan, the patient is placed on a table and low-dose x-rays beams are passed

from below through the area being measured. The exam is non-invasive, typically takes 10–15minutes (including preparation of the patient) and gives no pain or discomfort. The amount ofradiation the patient absorbs is low, about one tenth of the amount of a chest x-ray exam8.The x-rays are acquired by a detector in the instrument's arm. The instrumentation transduces thesignals received by the detector into an image of the skeleton, and analyses the quantity of bonein the skeleton7. Data can be obtained as bone mineral content (BMC) in g/cm or areal BMD) ing/cm2.

DEXA measurement sites can be central or peripheral. The lumbar spine (posterior-anteriorand lateral projections), and the proximal femur are defined central sites. The forearm and thecalcaneus are defined peripheral sites. DEXA can be also performed in “total body” mode, in orderto achieve overall and regional assessment of body composition in terms of fat, lean mass andbone9, but this is not appropriate to diagnose osteoporosis10.

The images acquired are processed by a dedicated software (generally provided together withthe scanner) that allows the clinician to make his/her assessment. After acquisition, the clinician(radiologist) analyses the images and prepares the exam report. Test results are expressed as twoscores11.

• T-score: a number showing the individual’s BMD (expressed as standard deviations,SD) compared to the mean BMD of a “young normal” adult population of the samegender. There are different T-scores depending on which group of young adults wereused as reference (e.g. Caucasian women, Hispanic men)4

• Z-score: a number reflecting the individual’s BMD (expressed as standard deviations,SD) compared to the mean BMD of other people within the same age group. There arealso different Z-scores depending on the group used as a reference (e.g. group of thesame age, or group with the same age, race, gender and weight)12

4

According to WHO, there are four general diagnostic categories for women 4:

• Normal. A value of BMD within 1 standard deviation of the young adult referencemean (T-score >= -1).

• Low bone mass (osteopenia). A value of BMD 1 or more standard deviation below theyoung adult mean, but less than 2.5 standard deviations below this value (T-score <-1 and > -2.5).

• Osteoporosis. A value of BMD 2.5 or more standard deviations below the young adultmean (T-score <= -2.5).

• Severe osteoporosis (established osteoporosis). A value of BMD 2.5 or more standarddeviations below the young adult mean in the presence of one or more fragility fractures.

2.3 The alternative technologies and the choice of QUS as comparator At present, DEXA is the most widely accepted and used method for the identification of oste-

porosis . However, it has disadvantages the machines are not portable; it is impossible to assessbone architecture; some factors may interfere with the accuracy of measures (e.g. osteoarthritis,vertebral compression fractures13. Besides DEXA, other technologies can be used to measureBMD14-18: Quantitative Computed Tomography (QCT), Quantitative UltraSound (QUS), MagneticResonance Imaging (MRI), Radiographic Absorptiometry (RA), Single and Double PhotonAbsorptiometry (SPA and DPA). A brief description of each is in Table 2.1.

Given the range of possible alternatives to DEXA as comparators for our assessment, and theneed to restrict the choice to only one, we decided to base our choice on the contextual eviden-ce provided by the available administrative data on the most used procedures at ambulatory levelin 2007 in the Regions which make up the Italian National Service (INHS). Data provided by 14Regions out of 21 (see chapter 6 for a detailed description of materials and methods) showed thatthe total highest volumes of procedures are those of DEXA (94%), followed by QCT (3,1%) andQUS (2,7%) (see chapter 6, table 6.2 ).

The most used technology in the Regions was undoubtedly DEXA and there was a small dif-ference in utilisation between QCT and QUS. A more detailed analysis of the regional data sho-wed that 8 out 12 regions used more QUS than QCT (Friuli Venezia Giulia and Valle d’Aosta, per-formed only DEXA). Eighty five percent of all QCT procedures were performed in 4 Regions out of12: Liguria, Lombardia, Piemonte and Puglia. So if we do not consider the total volume of activi-ty as a whole, and focus on the data provided by each single region, QUS results to be the devi-ce that most Regions use as an alternative to DEXA: Basilicata, P.A. Bolzano, Emilia Romagna,Lazio, Marche, Sardegna, Toscana, Veneto.

QUS is sometimes marketed as a safer, cheaper and more logistically flexible alternative toDEXA which can be used to carry out “walk in” assessments19. Such a pitch might induce a higherdemand for bone density assessment and fuel demand for interventions to prevent or minimisethe impact of osteoporosis.

Table 2.1: Alternative technologies for bone measurements (in alphabetical order) (WHO 2000).

We thus identified QUS as the alternative technology to evaluate the cost effectiveness ofDEXA (see chapter 7). Hereafter the report will focus on the comparison of DEXA with QUS.

2.4 DEXA and QUS: different measurements by different technologiesAccording to WHO, the gold standard for defining osteopenia or osteoporosis is DEXA of the

hip (femoral neck or total hip). However, many experts, including the International Society forClinical Densitometry, recommend using the lowest central DEXA T-score of postero-anterior lum-bar spine, femoral neck, or total hip for the identification of osteporosis20. DEXA measurements ofBMD at other sites (including the trochanter, Ward triangle, lateral lumbar spine, other forearmregions, heel, or total body) or with other technologies (calcaneal QUS, peripheral DEXA, QCT,SPA, DPA, or MRI) may be useful for assessing risk of fracture, but they are not recommended foridentification of osteoporosis21.

DEXA machines are large and seem to require a relatively high degree of operator skill (a trai-ned certified radiology technician) in order to position the anatomical target area correctly. DEXAmay be not readily available in all clinical settings and requires a dedicated examination room. Dueto these logistic limitations, the interest in QUS has increased and some countries have paid atten-tion to QUS as an emerging technology22. QUS is described as a portable office-based technolo-

5

Magnetic Resonance Imaging (MRI)

MRI does not provide direct information on bone density but about the quality of trabecular bone with highresolution. At present, MRI investigation of the skeleton remains a research procedure because of its highcosts and complexity (see Appendix 1 for details).

Radiographic Absorptiometry (RA)

RA was developed to determine BMD with plain x-ray and standard radiographic units. Comparisonsbetween the density of the bone and a reference aluminium wedge is performed by digitalised imagesprocessed with a computer (see Appendix 1 for details).

Single and Double Photon Absorptiometry (SPA and DPA)

SPA allows the scan of radius or calcaneus with a single-energy photon beam. The BMD is then calculat-ed from the images. SPA can be only performed peripherally since it is affected by the amount of soft tis-sue in the region of interest. DPA uses a double-energy photon beam and allows the determination ofBMD in spine, proximal femur and in the whole body. Both SPA and DPA use radioactive sources thatmust be replaced periodically (see Appendix 1 for details).

Quantitative Computed Tomography (QCT)

QCT uses x-rays computed tomography to perform volumetric measurements of BMD. However, it isexpensive and time consuming and exposes patients to marked amounts of radiation (see Appendix 1 fordetails).

Quantitative Ultrasound (QUS)

QUS uses sound waves to assess some bone features. The parameters measured are QUS-specific andcan be used to assess BMD. However QUS can be only used in peripheral sites (see Appendix 1 fordetails).

gy requiring less costly instruments, less space and less operator skill than DEXA and may beco-me more common in primary care offices and smaller clinics23. Moreover, QUS does not exposepatients to any ionising radiation. This statement is not supported by the findings of our survey(see chapter 6). The majority of our centres involve a physician as a QUS operator. This, as weshall see, is the main reason for the higher overall cost of the basic QUS procedure compared tothat of the basic DEXA procedure.

As briefly described in paragraph 2.2, DEXA measurements are expressed as T scores.Parameters generally measured by commercial QUS devices are the speed of sound (SOS) repor-ted as m/s and the broadband ultrasound attenuation (BUA) reported as dB/MHz. In the case ofosteopenia (or osteoporosis), reductions in bone density and trabecular number occur and reduc-tion in SOS and BUA can be observed22. BMD can be estimated using SOS and BUA but measu-res cannot be directly compared with DEXA parameters since the measured properties of the boneare different22. QUS measurements may be used to calculate a QUS T-score (which is differentfrom the DEXA T-score), which may vary with the QUS device used24. Other parameters measu-rable with QUS are the stiffness index (SI) and the quantitative ultrasound index (QUI) and Ad –SOS (amplitude dependent speed of sound) measured at the phalanges. SI and QUI are general-ly calculated combining SOS and BUA in order to achieve a single clinical measure with a lowerprecision error25-26. Correlation between QUS T-score and DEXA T-score is matter of research forseveral groups3,27. The problems are not only related to the differences in the devices, for bothDEXA and QUS, but also to the different bone characteristics measured and to the different mea-surement sites (central sites for DEXA, peripheral sites for QUS).

6

7

3. The marketing status of DEXA and QUS and currentreimbursement arrangements

3.1 DEXA We could find evidence that at least four companies sold DEXA machines in Italy: HOLOGIC

(two types: Discovery and Explorer), GE LUNAR, UBIS and NORLAND. Our survey (see chapter 6)involving a purposive sample of Italian centres, highlighted that among the 12 centres providingDEXA exam, 4 of them had 2 devices. Manufacturers of Dexa devices in Italy are at least 4: HOLO-GIC, GE LUNAR, UBIS, NORLAND. Ten of the above 12 centres (83%) had just HOLOGIC devices;one centre (8.3%) had only GE Lunar (8.3%) while in just one centre both manufacturers devi-ces were present.

In table 3.1 we summarise the characteristics of the devices reported by respondent centres(manufacturer, type, CE year mark).

The reimbursement arrangements for DEXA vary according to the number of screened skele-tal sites. The Ministerial decree on the INHS outpatient tariffs (“Prestazioni di assistenza speciali-stica ambulatoriale erogabili nell’ambito del Servizio Sanitario Nazionale e relative tariffe” – 22nd

July 1996) states that reimbursement for a total body DEXA is € 43,38, while a lumbar, femoraland ultra distal DEXA measurement is € 31,50.

Regions are allowed to increase or decrease the national reimbursement arrangements: in thecase of DEXA, Friuli Venezia Giulia, Trento P.A, and Piemonte have amended the national tariffs1

(see table 3.2).

3.2 QUSOur survey (see chapter 6) involving a purposive sample of Italian centres, highlighted that

among the 10 centres providing QUS exam, only 1 of them had 2 devices. Manufacturers of QUSdevices in Italy are at least 3: IGEA, HOLOGIC and GE LUNAR. Only 7 of the above 10 centres(70%) had explicitly reported the equipment type (3 centres GE LUNAR; 3 centres IGEA; 1 cen-tre HOLOGIC).

Table 1 summarises the characteristics of the devices reported by the respondent centres

1ASSR standing research programme “Programma di ricerca ed analisi del contenuto del livello essenzialedi assistenza specialistica ambulatoriale”, 2003-2005.

8

Table 3.1 Types of equipment in use in the responding centres by CE mark as declared by producers

Source: Agenas survey and information provided by manufacturers.

Table 3.2 National and regional reimbursement tariffs for outpatients procedures

Technology Manufacturer System name(questionnaires) CE mark

DEXA

GE Medical Systems Lunar Prodigy 1999

Hologic, Inc. Discovery 2003

Hologic, Inc. Explorer 2004

Hologic, Inc. QDR 4500 Off the market

Hologic, Inc. Delphi Off the market

QUS

GE Medical Systems Lunar Achilles InSight 2002

GE Medical Systems Lunar Achilles Plus Off the market

Hologic, Inc. Sahara 1997

IGEA S.p.A. DBM Sonic Bone Profiler 1998

IGEA S.p.A. 1200 DB Off the market

Ministry of Health reimbursements

Region Procedurecode Procedure Tariff in Euro

Italy 88.99.1 SINGLE OR DOUBLE PHOTON ABSORPTIOMETRY - RADIUS OR CALCANEOUS € 21,17

Italy 88.99.2 X-RAY BONE DENSITOMETRY - LUMBAR, FEMORAL, PERIPHERAL € 31,50

Italy 88.99.3 X-RAY BONE DENSITOMETRY - TOTAL BODY € 43,38

Italy 88.99.4 CT BONE DENSITOMETRY – LUMBAR € 76,95

Italy 88.99.5 ULTRASONIC BONE DENSITOMETRY € 17,56

Regions’ amended reimbursements

Region Procedurecode Procedure Tariff in Euro

P.A. Trento 88.99.3 X-RAY BONE DENSITOMETRY - TOTAL BODY € 45,50

P.A. Trento 88.99.4 CT BONE DENSITOMETRY – LUMBAR € 80,70

Friuli Venezia Giulia 88.99.1 SINGLE OR DOUBLE PHOTON ABSORPTIOMETRY - RADIUS OR CALCANEOUS € 25,30

Friuli Venezia Giulia 88.99.2 X-RAY BONE DENSITOMETRY - LUMBAR, FEMORAL, PERIPHERAL € 37,80

Friuli Venezia Giulia 88.99.3 X-RAY BONE DENSITOMETRY - TOTAL BODY € 52,00

Friuli Venezia Giulia 88.99.4 CT BONE DENSITOMETRY – LUMBAR € 92,30

Friuli Venezia Giulia 88.99.5 ULTRASONIC BONE DENSITOMETRY € 21,10

Note: This table was translated from Italian. Code 88.99.1 is generally performed by SPA or DPA; Code88.99.2 and 88.99.3 is generally performed by DEXA; Code 88.99.4 is generally performed by QCT; andCode 88.99.5 is generally performed by QUS.

9

4. Report’s objectives: policy and research question

Our Policy Question (PQ) aimed at understanding if DEXA, which is widely recognised by thescientific community as the gold standard for the identification of osteoporosis, is also the mostcost effective technique for the identification of osteoporosis in postmenopausal women, whencompared with the other most commonly used alternative in Italy.

To respond to this PQ we answered the following research questions:

• What is the scientific evidence on the accuracy and safety of DEXA versus the mostwidespread alternative in our context for the identification of primary osteoporosis inour target group?

• Which are the most widely used technologies for the identification of osteoporosis inItaly, how they are used and what are their costs?

• Given the context specific data collected, which is the most cost effective technologyfor the identification of primary osteoporosis in the target population?

The current report has thus four main objectives:

1) to understand which are the most widespread technologies for the identification ofosteoporosis in the target population in Italy.

2) to identify, retrieve, update and evaluate the evidence on the chosen technologies,considering different outcomes such as accuracy in the identification of osteporosissafety and economic impact.

3) to describe the acquirement and use of the technology and its alternative(s) in theItalian context.

4) to perform a cost-effectiveness analysis comparing the gold standard, DEXA, with themost widespread alternatives in the target population.

In chapter 5 (Systematic review), we describe methods and results of the systematic review(objective 2). In chapter 6 (context-specific data collection and analysis) we illustrate the analysisof administrative data performed and the materials, methods and results of the survey we carriedout (objective 1 and 3). In chapter 7 the cost effectiveness evaluation methods and results aredescribed and discussed (objective 4).

11

5. Systematic review

5.1 Methods We conducted searches on the following databases: PubMed, Medline, Embase, Cochrane

Library (CL). The time range was 1998-2008 (see Appendix 2 for the search strategy), except forthe search done on the CL were the time range was 1998-1/2009.

Assessment of the quality of studies was carried out using the QUADAS - Quality Assessmentof Diagnostic Accuracy Studies - checklist (see Appendix 3 ). We did not use a scoring system forthe evidence. A QUADAS form was compiled for each study. .

Inclusion criteria

We selected all the comparative studies assessing the accuracy of DEXA versus QUS for theidentification of primary osteoporosis in Caucasian, healthy postmenopausal women. We includedprimary or synthesis studies on Caucasian postmenopausal women comparing the performance ofDEXA vs QUS for identification of primary osteoporosis. Participating women had to be healthy andnot undergoing either dietary, lifestyle or treatment interventions. Data were extracted from stu-dies with larger populations, provided they reported data broken down by gender, health andintervention status.

5.2 Results A total of 184 primary studies were identified by the search (61 studies from Embase, 45 from

CRD DARE, 39 from PubMed and CL each). Seven out of 184 studies met the inclusion criteria (seeflow chart in figure 5.1). A list of all excluded studies with reasons for exclusion is provided inAppendix 4.

5.2.1 Safety

Radiation exposure from DEXA scans is relatively low. Moreover, the radiation exposure can beminimised during the procedure, for example, by choosing fast scan speeds, and by positioningthe patient to avoid to repeating the scan. Radiation exposure associated with DEXA can be com-pared with the natural background radiation exposure (which comes from terrestrial sources suchas soil and rocks, and from extraterrestrial sources such as cosmic rays). In the USA the averageannual background radiation exposure is about 8 Sv/d (micro-Sievert per day)27.

As a comparison, radiation exposure associated with a DEXA scan is generally less than 1 dayof natural background radiation exposure. It has been estimated that one lumbar spine DEXA scanmight increase the risk of cancer by less than 0.0001%28. Given that a dose of ionising radiationis absorbed, other technologies with the same diagnostic capability could be considered as analternative.

12

5.2.2 Accuracy of DEXA versus QUS in the identification of osteoporosis

The description and our evalutation of each of the 7 included studies is at Appendix 5. Ourquality assessment for each studiy is available at Appendix 6. In general we found no randomisedcomparisons but were able to identify several cohort studies3,19,29,30,31,32 directly comparing thetwo techniques and one cross sectional study33 from Iran. Reporting quality was variable as werethe populations and settings of the assessments.

The best designed study32 was carried out on 7,598 French women aged more than 75. It con-cluded that in this age group both techniques were of equivalent accuracy. Although the study wascarried out in agreement with strict quality control procedures (assessment of interoperator varia-bility and cross-calibration) such strict methods may not be applicable to routine clinical settings.

The other largest studies3,29 were both carried out in UK in the age group of interest to us.The studies concluded that both techniques had similar, high, sensitivities (around 96% whencompared to total hip DEXA) but different specificities. Clowes et al. report that DEXA had a spe-cificity of up to 72% whereas the best QUS specificity (Metra BUA) was 54% compared to totalhip DEXA. In this well reported study there were marked differences in the devices-specific abso-lutes values. The positive and negative predictive values were device-specific and site-specific.This emphasises that it is important to consider these factors in the use of peripheral device in cli-nical practice.

The remaining 3 studies (2 from Turkey and 1 from Iran) were much smaller and reached simi-lar conclusions. The conclusions of the cohort study by Dane et al are limited by using calcaneusQUS. All calcaneus devices use the T score WHO thresholds to compare with on DXA lumbar spineand have low sensitivity and specificity (as the authors conclude).

The other Turkish study, by Jemalmaz and colleagues uses a small sample size to find a cut-off value for QUS. No data about fragility fracture in the screened population are reported, but themajority of women were overweight and obese. This may reduce the precision and accuracy ofthe measurement by both DEXA and QUS.

In the cross sectional study by Larjiani and colleagues the use of T score WHO thresholds incalcaneal QUS devices reduces the pratical application in a large population. However, develop-ment of specific cut-off values may optimise the usefulness of QUS in screening studies.

The small US cohort study by MacLaughlin and colleagues was a community screening studyon 97 postmenopausal women. The women were subjected to a QUS-DEXA screening sequenceto test the possibility of identifying cases of osteoporosis with QUS left heel in a general popula-tion for further assessment with DEXA. Comparison of performance was achieved using a Stiffnessindex (SI) – derived T score. The small sample included postmenopausal women from 55 years orolder of mixed ethnicity but the normative reference curve for the calcaneal QUS used in study isconstructed only from Caucasian women). No data about DEXA devices used are reported nor theprecision and sensitivity of QUS devices. Therefore, no conclusions about the use of QUS in thescreening program can be drawn. The authors conclude that QUS could be useful in detectingcases of osteoporosis, but such conclusions should be regarded with caution in the light of thesmall population and high risk of false positives.

Perhaps the most radical conclusions are those of Frost and colleagues, who suggested achange in WHO definitions for osteoporosis and osteopenia (i.e. a lowering of the diagnostic thre-sholds). While this may make the performance of the two techniques comparable, it would havethe effect of significantly increasing the size of the osteoporotic population. Two different QUSdevices were used in the study, but data on cross-calibration, accuracy and precision of the useddevices are not reported. These data are particularly important to evaluate the precision of ultra-sound. T and Z scores were calculated using local reference data constructed with a relativelysmall sample size. In addition WHO criteria constructed with lumbar spine DXA (X-ray) data, can-not be applicable to a QUS device based on ultrasound. However we agree with the Authors’ con-clusion that “.. at present criteria for the identification of osteoporosis and strategies for usingQUS for fracture risk assessment need to be developed before QUS at the calcaneus can be usedmore effectively in the clinical practice”.

5.3 ConclusionsAlthough portable and less expensive according to best available evidence, QUS has a compa-

rable sensitivity but a lower specificity than DEXA. However, QUS specificity increases with age32.Any suggestion that QUS can be used for stratifying the population into risk groups in a commu-nity screening setting should be subject to an economic evaluation.

13

Figure 5.1

Flow chart

14

Identified by the search

N=184

Items excluded because not reporting original data (e.g.

editorial or news article) or had no author

Original comparative Studies assessing

N=182


N=182


N=121

Studies excluded because not carried out on

humans

Studies excluded because no comparisons of

DEXA with QUS were made.

Studies excluded because not on a population with primary

osteoporosis, postmenopausal caucasian women or no

sensitivity and specificity data for DEXA or QUS were reported.


DEXA vs QUS

Available studies for the comparative assessment of the effectiveness of DEXA vs QUS

N=7

15

6. Context specific analysis

We analysed administrative data and collected primary contextual information to understandwhich are the two most widespread technologies for the identification of osteoporosis in the tar-get population and their specific characteristics.

6.1 Materials and methods

Administrative data

For the collection and analysis of existing administrative data, we decided to follow two diffe-rent paths (as inpatients and outpatients) to have a more complete picture of the volumes ofDEXA, QUS, QCT and RA activity for primary osteoporosis.

To quantify the volumes of DEXA, QUS, QCT and RA procedures for primary osteoporosis per-formed at inpatient level, we used the Italian hospital discharge records (SDO) database. This isthe main reliable and validated source of data from the Italian Ministry of Health. The two classi-fication codes that were identified by our clinical experts as indicating DEXA, QUS, QCT and RAprocedures were: ICD9-CM 88.79 “Other diagnostic ultrasound (ultrasonography of: multiple sites,non gravid uterus, and total body)” and ICD9-CM 88.98 “Bone mineral density studies (dual pho-ton absorptiometry; quantitative computed tomography studies; radiographic densitometry; singlephoton absorptiometry)”. The two codes were considered to be too comprehensive: data could notbe disaggregated by technique and, as a result, aggregate information from this database was notuseful. In addition, the majority of DEXA, QUS, QCT and RA procedures for the identification ofprimary osteoporosis are supposed to be at outpatient level on asymptomatic women who are rou-tinely screened for osteoporosis. In the light of these considerations we decided not to analysethe SDO database and to focus exclusively on outpatient data.

Data on DEXA, QUS, QCT and RA volumes on outpatient level of care were formally requestedfrom each Region since no national validated and systematic database is available for this level ofcare. We requested data for volumes of outpatient procedures for each technology carried outduring 2006-2007. We asked for a minimum set of variables including: number of procedures by5 Italian outpatients codes2, number of procedures by gender, age and typology of providers. Wehad complete feedback from 14 Regions out of 21: Basilicata, P.A. Bolzano, Emilia Romagna, FriuliVenezia Giulia, Lazio, Liguria, Lombardia, Marche, Piemonte, Puglia, Sardegna, Toscana Valled’Aosta, Veneto. The Campania Region reported that an outpatient database did not yet exist (seeFig. 6.1). Sicilia, Calabria, Molise, Abruzzo, Umbria, and Trento P.A. did not respond to our request.‘

2See table 3.2, chapter 3, for a list of the procedures’ codes.

16

Fig.6.1 Availability of outpatients administrative data in Italian Regions, responding and nonresponding Regions.

Survey

On the basis of the administrative data collected from the 14 responding Regions (in red,fig.6.1) we extracted a purposive sample3 of centres performing only DEXA, both QUS and DEXAand only QUS. We thus aggregated the regional data into the three typologies of providers. Wethen selected the provider with the highest volumes of procedures for each group. We excludedcentres with less than 10 procedures by type of technology.

Regions which sent the requested data (14)

Regions with no data (1)

Non responding Regions (6)

3 Purposive sampling is a non probabilistic sample chosen on the basis of what the researcher thinks wouldbe appropriate for the study

We defined the sample of 36 centres as per Table 6.1.

Table 6.1: Number of selected centres broken down by region and procedure

Thirteen centres performed DEXA, 12 performed DEXA and QUS and 11 performed only QUS.A structured questionnaire (see Appendix 7) was sent at the end of April 2009 to all selected cen-tres, which were asked to return the filled questionnaire within three weeks. The instrument hadbeen tested in January 2009 at Policlinico Umberto I (in Rome), one of the providers with thehighest volume of procedures in 2008.

The questionnaire was divided into three parts.

• Characteristics of the procedures (total volumes, number of patients, number ofexams due to primary osteoporosis, type of exam, volumes broken down per age) forthe year 2008

• Characteristics of procedures for postmenopausal women (total number of womenand postmenopausal women, reasons, results) for the year 2008

• Organisational and economical aspects of the procedures (time of health personnel,quantity of consumables, type and price of devices etc. ) for the year 2008.

RegionNumber of centres performing:

Total centresDEXA+QUS DEXA QUS

Basilicata 1 1 - 2

Emilia Romagna 1 1 1 3

Friuli Venezia Giulia - 1 - 1

Lazio 1 1 1 3

Liguria 1 1 1 3

Lombardia 1 1 1 3

Marche 1 1 1 3

P.A. Bolzano 1 - 1 2

Piemonte 1 1 1 3

Puglia 1 1 1 3

Sardegna 1 1 1 3

Toscana 1 1 1 3

Valle d'Aosta - 1 - 1

Veneto 1 1 1 3

Total centres 12 13 11 36

17

6.2 Results

6.2.1 Analysis of administrative data

The data show that the most used technology is DEXA (94%) and this seems to be used inan appropriate manner (Mauck KF 2006) as Local DEXA (i.e. lumbar and femoral bone, wrist, fore-arm, vertebral) has the highest volumes (90%), while the remaining 4% are Total DEXA procedu-res. Regions performing more local DEXA are Valle d’Aosta (99%) and Friulia Venezia Giulia(100%). In the latter Region the core framework of care includes just local DEXA, which is theonly one, among the technologies under analysis, that can be reimbursed. Data also show thatthe Marche Region does not perform Local DEXA at all and has low volumes of Total DEXA (27.4%)and that the most frequent procedure was QUS (58,7%) (see table 6.2.).

Table 6.2: volumes of procedures for DEXA, RA, QTC, QUS in outpatient care in 14 ItalianRegions in 2007

Source: Agenas analysis based on outpatient administrative data from 14 Regions.

Key: DEXA_L= local DEXA (central or peripheral), DEXA_T=Total Dexa, QCT= Quantitative ComputedTomography; SPA and DPA: Single and Double Photon Absorptiometry. P.A.= Autonomous Province.

Region

PercentagesAbsolute valuesTotalDEXA_L DEXA_T SPA and DPA QTC QUS Total

BASILICATA 83.9 - - 6.0 10.0 100.0 9,358

P.A. BOLZANO 75.8 0.3 0.1 3.7 20.0 100.0 8272

EMILIA ROMAGNA 90.0 5.8 2.2 0.3 1.8 100.0 93,965

FRIULI V. GIULIA 100.0 - - - - 100.0 13,618

LAZIO 87.8 8.3 0.2 0.9 2.7 100.0 207,778

LIGURIA 91.2 4.5 0.6 1.9 1.8 100.0 64,817

LOMBARDIA 95.3 1.0 0.1 2.8 0.8 100.0 212,496

MARCHE - 27.4 - 13.9 58.7 100.0 3,085

PIEMONTE 81.8 5.7 0.3 9.6 2.6 100.0 94,030

PUGLIA 76.7 0.1 0.2 14.2 8.8 100.0 72,353

SARDEGNA 98.4 0.8 0.5 0.1 0.2 100.0 29,506

TOSCANA 93.0 5.4 0.6 0.1 0.9 100.0 119,791

VALLE D'AOSTA 99.3 0.7 - - - 100.0 1,384

VENETO 94.6 0.0 0.0 1.7 3.7 100.0 58,019

Total % 89.6 4.1 0.5 3.1 2.7 100.0

Total Absolute values 885,434 40,738 4,466 31,057 26,777 988,472

18

The data on volumes of outpatient procedures for the identification of osteoporosis brokendown by gender in each Region, show that 96% of procedures are performed on women in almostall Regions, except for Emilia Romagna and Marche where the percentage of men undergoingthese procedures almost doubles (from an average of 4% to 9%). The central Italian region ofMarche is also an outlier for the high proportion of QUS procedures carried out, although overallvolumes of acivity are low (Table 6.2 and Figures 6.2 and 6.3). We cannot account for this finding.

Fig. 6.2 Volumes of outpatient procedures for the identification of osteoporosis per gender byRegion §. Year 2007

Source: Agenas analysis based on outpatient administrative data from 14 Regions.§ Sardegna Region did not provide data broken down by age.

Focusing our analysis solely on women and on the type of the most used procedures in thispopulation group, the data confirm that local DEXA is the most used tecnique (93%), followed bytotal DEXA (4%) and QUS (2,8%) (Figure 6.3 and Tables 6.3).

19

0% 20% 40% 60% 80% 100%

All regions

VENETO

VALLE D'AOSTA

TOSCANA

SARDEGNA

PUGLIA

PIEMONTE

MARCHE

LOMBARDIA

LIGURIA

LAZIO

FRIULI V. GIULIA

EMILIA ROMAGNA

P.A. BOLZANO

BASILICATA

Females Males

Table 6.3: Volumes of outpatient procedures in women by age class, procedure and Region - year2007

Source: Agenas analysis based on outpatient administrative data from 14 Regions.Key: DEXA_L= local Dexa (central or peripheral), DEXA_T=Total Dexa.

(*) Sardegna Region did not provide data broken down by age.

Region

All ages Aged 50 or more

Percentages Absolutevalue Percentages Absolute

value

DEXA_L DEXA_T QUS Total Total DEXA_L DEXA_T QUS Total Total

BASILICATA 89.2 - 10.8 100.0 8,406 89.1 - 10.9 100.0 7,729

P.A. BOLZANO 78.7 0.3 21 100.0 7,593 79.2 0.1 20.7 100.0 6,952

EMILIA ROMAGNA 94.1 4 1.8 100.0 83,983 95.2 3.1 1.8 100.0 75,526

FRIULI V. GIULIA 100 - - 100.0 13,005 100.0 - - 100.0 11,767

LAZIO 88.9 8.4 2.7 100.0 195,703 89.0 8.2 2.8 100.0 178,487

LIGURIA 93.5 4.6 1.8 100.0 6,0517 93.6 4.6 1.9 100.0 56,808

LOMBARDIA 98.3 0.9 0.8 100.0 195,399 98.4 0.8 0.9 100.0 180,719

MARCHE - 33.3 66.7 100.0 2,422 - 33.3 66.7 100.0 1,614

PIEMONTE 90.8 6.4 2.8 100.0 80,611 91.0 6.3 2.8 100.0 74,354

PUGLIA 89.6 0.1 10.3 100.0 58,929 89.6 0.1 10.3 100.0 54,677

SARDEGNA 99.2 0.6 0.2 100.0 28,343 (*) (*) (*) (*)

TOSCANA 93.7 5.4 0.9 100.0 113,117 93.8 5.4 0.8 100.0 101,526

VALLE D'AOSTA 99.3 0.7 - 100.0 1,312 99.7 0.3 - 100.0 1,178

VENETO 96.3 0 3.7 100.0 54,345 96.3 0.0 3.7 100.0 52,746

All regions%

Absolute value

93.1 4.1 2.8 100.0 93.2 3.9 2.8 100.0

841,752 36,745 25,188 903,685 749,670 31,665 22,748 804,083

20

Fig. 6.3 Volumes of outpatient procedures in postmenopausal women for DEXA and QUS perRegion § – year 2007


Table 6.4 and Fig 6.4 show regional crude rates broken down by age groups. After the meno-pause, the rate of use of DEXA has greater variation between regions than in other age groups.This ranges from low use in the Marche Region (0.1%) to a maximum in Lazio (14%), while thereis a tendency to zero use in women older than 85 years. There is a broad growing trend in all agegroups until 65-75 years followed by a dramatic decrease.

In table 6.5 and Fig. 6.5 crude rates broken down by Region and age classes show a very lowrate of QUS use. Rates of QUS are slightly different rates by Regions (except P.A. Bolzano, Pugliaand Basilicata). There is a growing trend in the use of this procedure up to the 65-75 age group.Slightly fewer procedures are carried out in older age groups.

21

0% 20% 40% 60% 80% 100%

All regions

VENETO

VALLE D'AOSTA

TOSCANA

PUGLIA

PIEMONTE

MARCHE

LOMBARDIA

LIGURIA

LAZIO

FRIULI V. GIULIA

EMILIA ROMAGNA

P.A. BOLZANO

BASILICATA

DEXA QUS

Table 6.4: DEXA use (crude rates broken down by Region and age group per 100 postmenopau-sal women in 2007).


Fig 6.4 DEXA use (crude rates broken down by Region§ and age group per 100 postmenopausalwomen in 2007)


Region §Age groups

Total50-55 56-65 66-75 76-85

more than85

BASILICATA 4.5 6.7 8.5 3.6 0.6 5.8

P.A. BOLZANO 5.4 7.1 8.0 4.6 1.1 6.2

EMILIA ROMAGNA 8.0 10.2 9.8 4.1 0.5 7.7

FRIULI V. GIULIA 3.8 5.5 5.3 2.2 0.3 4.1

LAZIO 13.9 18.1 19.3 8.7 1.4 15.0

LIGURIA 11.0 16.1 19.1 8.4 1.1 13.3

LOMBARDIA 7.3 11.1 11.7 5.2 0.8 8.8

MARCHE 0.1 0.2 0.3 0.1 0.0 0.2

PIEMONTE 7.1 9.7 8.7 3.7 0.5 7.1

PUGLIA 4.7 7.0 9.4 3.9 0.5 6.3

TOSCANA 12.8 15.7 14.7 5.8 0.7 11.7

VALLE D'AOSTA 4.8 5.9 5.2 2.1 0.2 4.4

VENETO 0.9 2.8 11.8 5.2 0.6 5.1

Total Regions 7.5 10.4 11.8 5.1 0.7 8.6

22

0,0

2,0

4,0

6,0

8,0

10,0

12,0

14,0

16,0

18,0

20,0

50-55 56-65 66-75 76-85 more than 85

Age

Out

patie

ant D

EX

A c

rude

rat

e

BASILICATA P.A. BOLZANO EM ILIA ROM AGNA FRIULI V. GIULIA LAZIO

LIGURIA LOM BARDIA M ARCHE PIEM ONTE PUGLIA

TOSCANA VALLE D'AOSTA VENETO All regions

Table 6.5: QUS use (crude rates broken down by Region and age group per 100 postmenopau-sal women in 2007).


Fig. 6.5 QUS use (crude rates broken down by Region § and age group per 100 postmenopausalwomen in 2007)


Region §Age class

Total50-55 56-65 66-75 76-85

more than85

BASILICATA 0.7 0.9 0.9 0.5 0.0 0.7

P.A. BOLZANO 1.7 1.7 2.1 1.2 0.4 1.6

EMILIA ROMAGNA 0.2 0.2 0.2 0.1 0.0 0.1

FRIULI V. GIULIA - - - - - -

LAZIO 0.3 0.4 0.7 0.4 0.1 0.4

LIGURIA 0.2 0.2 0.4 0.2 0.0 0.2

LOMBARDIA 0.1 0.1 0.1 0.1 0.0 0.1

MARCHE 0.3 0.3 0.4 0.2 0.0 0.3

PIEMONTE 0.2 0.2 0.3 0.1 0.0 0.2

PUGLIA 0.6 0.8 1.0 0.5 0.1 0.7

TOSCANA 0.1 0.1 0.1 0.1 0.0 0.1

VALLE D'AOSTA - - - - - -

VENETO 0.0 0.1 0.4 0.3 0.1 0.2

All regions 0.2 0.3 0.4 0.2 0.0 0.3

23

0,0

0,5

1,0

1,5

2,0

2,5

50-55 56-65 66-75 76-85 more than 85

Age

Out

patie

ant Q

US

cru

de r

ate

BASILICATA P.A. BOLZANO EM ILIA ROM AGNA LAZIO

LIGURIA LOM BARDIA M ARCHE PIEM ONTE

PUGLIA TOSCANA VENETO All regions

Table 6.6 Standardised regional rates (adjusted by age) show an important variability amongRegions, both for DEXA and QUS (Variation Coefficient - VC for DEXA 55% and for QUS 104%)(table 6.6).

In the Lazio, Liguria and Tuscany Regions, DEXA has the highest rates of use, while for QUSthe P.A. of Bolzano has a significantly higher rate of use compared with the average rate.

Table 6.6: Age standardised rate and standard deviation of use by procedure broken down byRegion per 100 postmenopausal women.

Source: Agenas analysis based on outpatient administrative data from 14 Regions.§ Sardegna Region did not provide data broken down by age. °Standard Population: Italian population as reported by ISTAT 2007. *statistically significant 95%: means deviation divided by standard deviation more than 1.96 as an absolutevalue.

Fig. 6.6 is based on data from table 6.6 (QUS means deviation on the X axis and DEXA on theY axis). Each point in the figure represents a Region. The second quadrant of the Cartesian coor-dinate system indicates a highest use of DEXA compared with its alternative (QUS), while thefourth quadrant indicates the same for QUS compared with DEXA. The P.A. of Bolzano, Puglia andBasilicata seem to have a high use of QUS, while Lazio, Liguria and Tuscany show a high use ofDEXA.

Region §DEXA QUS

Standardised rate

Mean deviation

Standardised rate (°)

Mean deviation

BASILICATA 5.77 -1.62 0.71 0.28

P.A: BOLZANO 6.17 -1.23 1.61 1.18 *

EMILIA ROMAGNA 7.89 0.50 0.14 -0.29

FRIULI V. GIULIA 4.14 -3.25 - -

LAZIO 14.76 7.36 0.42 -0.01

LIGURIA 13.49 6.09 0.25 -0.18

LOMBARDIA 8.75 1.36 0.08 -0.36

MARCHE 0.16 -7.24 0.32 -0.11

PIEMONTE 7.19 -0.20 0.20 -0.23

PUGLIA 6.19 -1.20 0.71 0.28

TOSCANA 12.02 4.62 0.10 -0.33

VALLE D'AOSTA 4.43 -2.96 - -

VENETO 5.17 -2.23 0.20 -0.23

Means 7.39 0.43

Standard Deviation 4.05 0.45

VC 54.81 104.15

24

Fig. 6.6: Mean deviation of standardised rates for QUS and DEXA

Source: Agenas analysis based on outpatient administrative data from 14 Regions.

6.2.2 Survey

Seventeen out of thirty six selected centres responded to the survey, conducted by a question-naire. The returning rate was 47,2% but only 16 of the 17 centres were able to provide data (seeappendix 8).But response quality was variable as not all Italian centres providing outpatients pro-cedures collect the kind of data we asked (and usually have information just on the number ofprocedures). In our final analysis we included also the questionnaire from the pilot centre PoliclinicUmberto I in Rome. Our study is thus based on data from 16 centres: 6 questionnaires from thefirst group (DEXA), 4 questionnaires from the second group (QUS), and 6 questionnaires from theDEXA/QUS group The analysis of data shows no difference in the clinical data (volumes/patientsetc.) among the three type of centres, so data about DEXA are from all the centres providing theprocedure (12 centres) regardless of which of the above groups they belong to. The same appliesto QUS data (10 centres).

6.2.2.1 Characteristics of procedures and volumes of activity

Data from the responding centres (10/12 for DEXA and 7/10 for QUS) show that each patientundergoes an average of 1.5 DEXA exams per year, but with a high degree of variability from cen-tre to centre: from a minimum of 1 exam to a maximum of 2 (Median=1.46 for DEXA and 1 forQUS). QUS use is less variable and patients undergo an average of 1 exam per year. The group

25

P.A. BOLZANO

TOSCANA

PIEM ONTE

LOM BARDIA

VENETO

LIGURIA

EM ILIA ROM AGNA

M ARCHE

LAZIO

PUGLIABASILICATA

-8

0

8

-2 0 2

Means deviation for QUS

Mea

ns

dev

iati

on

fo

r D

EX

A

of patients with an age range 50-70 years is the one that use more DEXA/QUS procedures in theresponding centres (11/12 and 5/10).

Fig.6.7 Age of patients undergoing DEXA and QUS

Source: Agenas

The twelve responding centres performing DEXA or DEXA plus QUS carried out a total of93,862 procedures in 2008. The majority of DEXA examinations were carried out on an outpatientbasis (88%), while 7% were carried out as inpatients level and 5% in day hospitals. The sameapplies to QUS, as 92% of exams were carried in outpatients (see Table 6.7).

Table 6.7: Number of DEXA and QUS procedures by level of care, in 2008

Source: Agenas

Seven of the 12 centres reported the reasons for performing DEXA procedures: 80% are dueto suspected primary osteoporosis, 18% are done for secondary osteoporosis and 2% for other

Exam Inpatient Day Hospital Outpatient

DEXA 7% 5% 88%

QUS 4% 4% 92%

26

reasons. DEXA of the lumbar spine and femoral neck is the most used (94%; 12/12), while TotalBody DEXA and Vertebral morphometry represent just 2% and 3% of the total. DEXA of the fore-arm and wrist was not carried out by responding centres. Fifty-seven percent of QUS procedureswere carried out on the phalanges and 43% on the heel (Table 6.8).

Table 6.8: Type of DEXA or QUS carried out in 2008

Source: Agenas

In 7 of 12 responding centres, 80% of DEXA exams were undertaken for suspected primaryosteoporosis. QUS was undertaken for the same purpose (84%). (Table 6.9)

Table 6.9: Why DEXA and QUS exams were carried out

Source: Agenas

6.2.2.2 Characteristics of procedures for postmenopausal women

In 10 of the 12 responding centres 76,5% of women undergoing DEXA in 2008 were postme-nopausal; 79% of women undergoing QUS in 6 of the 10 selected centres were postmenopausal.In 9 of the 12 responding centres DEXA was performed for prevention (44%), bone pain (35%),other disease (14%) and fractures (6%). In 4 of the 12 responding centres QUS was carried outmainly for prevention (50%) bone pain (29%), other disease (15%) or fractures (6%) (Table 6.10).

Table 6.10: Reasons for DEXA and QUS use in postmenopausal women ( 2008)

Source: Agenas

Exam Fracture Bone pain Prevention Other disease

DEXA 6% 35% 44% 14%

QUS 6% 29% 50% 15%

ExamPrimary

osteoporosis Secondary

osteoporosisOther

DEXA 80% 18% 2%

QUS 84% 13% 3%

Exam Lumbar-femoral Forearm and wrist Total bodyVertebral

morphometry

DEXA 94% - 2% 3%

Exam Phalanges Heel Other

QUS 57% 43% -

27

6.2.2.3 Organisational and economic aspects of technologies and procedures

Resource use of DEXA and QUS procedures

Responding centres were asked to report the average time (in minutes) spent by four typesof health personnel: physicians, radiologists, nurses, specialising physicians, administrative offi-cers. Procedures were divided into two phases, and we asked how much time each professionalspent in each phase as follows:

• Consultation: time spent in meeting the patient and discussing procedure and futureresults. In the case of administrative staff this results in booking the procedure forthe patient and in the time spent for the payment of the ticket (a state fee levied onnon-exempt persons).

• Procedure: the time for the preparation of the patient and for the exam to be done.

Health personnel and time for a DEXA examination

Eleven of the responding centres (92%) returned data on time spent by the different profes-sionals on a standard DEXA procedure (Table 6.11). Calculation of the average time spent hasbeen done for each type of professional involved and, when feasible, we also highlighted the mini-mum and maximum time and the mode.

Data from 7 of the 12 centres show that physician average time for one single DEXA exam isestimated to be around 17 minutes. The mode value shows a time of 15 minutes. The radiologistaverage time for one single DEXA exam was estimated to be around 17 minutes (4 minutes forconsultation and 13 for the exam (11/12) while the mode was 15 minutes. A specialist registrarphysician (i.e. a trainee specialist) (in 2 out of 12 centres) spends an average time of 21 minutes(8 minutes for consultation and 13 for the exam). The nurses’ average time spent is about 19minutes (4 minutes in consultation and 15 during the exam). Average time spent by administrati-ve staff (9/12) was 15 minutes (9 minutes for booking and payment and 15 for handling the exa-m’s results), the mode was 20 minutes. Centres indicating “specialist registrar” “other personnel”and “nurses” as professionals involved in the procedure, were only 2 out of 12 for specialist regi-strar doctor, 1 out of 12 for other personnel and 3 out of 12 for nurses. For this reason we did notinclude the average time spent by the three above professional groups in our final calculation ofthe average time by professional category.

28

Table. 6.11: Average time (in minutes) spent on a single DEXA procedure

*Including booking time and ticket payment in case of administrative staff.

Consultation* Exam*

Physician

Average 7 10

min-max (2-10) (5-15)

Mode 10 5

Radiologist

media 4 13

min-max (2-10) (7-20)

Mode 5 10

Specialist registrar physician

Average 8 13

min-max (5-10) (10-15)

Mode

Nurse

Average 4 15

min-max (2-15) (5-15)

Mode 5 15

Other personel

Average 5 15

min-max (5-15) (5-15)

Mode

Administrative

Average 9 6

min-max (5-15) (2-10)

Mode 10 10

29

Health personnel and time for a QUS exam

Seven out of 10 centres returned data on time spent by the different professionals on a stan-dard QUS procedure (Table 6.12). As in the case of DEXA, calculation of the average time spenthas been carried out for each type of professional involved and, when feasible, we also highlightedthe minimum and maximum time and the mode.

Physician average time for a single QUS exam was estimated to be around 19 minutes: 10minutes for consultation and 9 for the exam (7/10 centres). The mode value shows a time of 15minutes (5 for consultation and 10 for the exam). The radiologist average time (3/10 centres) forone single QUS exam was estimated around 20 minutes (10 minutes for consultation and 10 minu-tes for the exam), while the mode is 20 minutes. The nurse’s average time (1/10 centres) wasabout 15 minutes spent just on the exam. Average time for administrative staff (2/10) was 10minutes (5 minutes for booking and payment and 5 for handling the exam’s results).

Unlike in DEXA, in the case of a QUS procedure the health operator who spends more time inperforming the procedures is the physician (7/10 centres). Administrative personnel do not spendmuch time on a QUS procedure (2/10 centres).

Table. 6.12: Average time in minutes spent on a single QUS procedure

*This includes booking time and ticket payment in case of administrative staff.

Consultation* Exam*

Physician

Mean 10 9

min-max (5-15) (5-10)

Mode 5 10

Radiologist

Mean 10 10

min-max (10) (10)

Mode 10 10

Nurse

Mean 15

min-max

Mode

Administrative

Mean 5 5

min-max (5) (5)

Mode 5 5

30

Consumables for DEXA procedures

Consumables (table 6.13) include: stationery, paper rolls for the bed, disinfectant and gloves.

Table 6.13: Consumables for a single DEXA procedure

Source:Agenas

Consumables for QUS procedures

Consumables (Table 6.14) used for a single QUS include: stationery, paper bed rolls, disinfec-tant and gel for ultrasound scanning.

Table 6.14: Consumable for a single QUS procedure

Source:Agenas

Number of Centres

Printer toner 2/10

Printer cartridges 4/10

Paper 3/10

Files 1/10

Paper bed rolls 2/10

Disinfectant bottle 2/10

Gel for ultrasound scanning 2/10

Number of centres

Toner 8/12

Printer cartridges 8/12

Paper 10/12

Envelopes 3/12

Paper rollers 11/12

Disinfectant 8/12

Gloves 6/12

31

DEXA Equipment

All 12 responding centres reported data on the type of equipment used for a DEXA exam. DEXAmachines had been bought between 1999 and 2008; HOLOGIC was present in 10 centres, GELUNAR was present in only one and one centre declared the use of equipment from both manufac-turers (HOLOGIC and GE LUNAR).

QUS Equipment

Seventy percent (7/10) of the responding centres contributed data on the type of equipmentused for a QUS exam. Equipment had been bought between 1996 and 2004. Three centres had aGE LUNAR, three centres had an IGEA and one centre had the HOLOGIC.

32

33

7. Cost effectiveness analysis

7.1 Methods Given the high number of procedures performed in the INHS, and the high prevalence of

osteoporosis, there are a host of potential benefits accruing from identifying the most cost-effec-tive procedure to quantify bone fragility. Thus our cost-effectiveness analysis aimed to comparethe gold standard, DEXA, with QUS in our target population to evaluate the Incremental CostEffectiveness Ratio (ICER) of the two technologies. We adopted the point of view of the INHS.

For the efficacy in the identification of osteporosis we considered as outcomes the number ofpatients with osteoporosis identified by DEXA compared with QUS. The study by Clowes (2006),provided good quality evidence on the efficacy of DEXA and QUS in our target population and weused its data.

To identify, enumerate, measure and value (in 2008 Euros) all resources (average times byhealth personnel, materials, etc.) used to perform DEXA and QUS procedures in the Italian con-text (the 14 responding Regions) we used our survey results as a source of national data.

Our sensitivity analysis was based on two variables: one economic and the other epidemiolo-gical. Among all the economic variables, we focused on cost of equipment since it showed to havethe highest influence on the total costs. For the epidemiological variable we considered the pre-valence of osteporosis estimated by Clowes 2006 which was 9.8% and the 20% of osteporosisprevalence in Italian women as reported by the ESOPO study (2004).

7.2 Efficacy of DEXA and QUS for the identification of osteporosis in thetarget population

We used the prevalence of osteoporosis, the sensibility and specificity reported by Clowes etal. 2006 to estimate the percentage of true positives in the study population, which was not expli-citly reported in the above study. We assumed that prevalence and the true positives in Clowes etal population could be the same in our Italian target population (see Table 7.1a) Table 7.1b reportsthe data used for the sensitivity analysis based on the Esopo study prevalence of osteoporosis.

Table 7.1a: Data and sources of data for the economic evaluation

SOURCES OF DATA DATA VALUE

ISTAT 2008 Italian female population >50 years 12,614,484

Clowes et al. 2006

Prevalence 9.8%

Achilles BUA (T score -0.81)• Sensibility• Specificity• Estimated true positive

96%55%

9.4%DBM Ad SOS (T score -1.96)• Sensibility• Specificity • Estimated true positive

96%33%

9.4%

Table 7.1b: Data and sources of data for the economic evaluation (prevalence 20%)

7.3. Costs analysis for DEXA and QUS procedures We used standard costing for the assessment of the elements of cost. The costs attributed to

the various elements used for a single DEXA and a single QUS procedure were derived as an ave-rage of the declared values by responding centres or, when incomplete, were derived from mar-ket prices. All costs were inclusive of VAT (20%).

The mean personnel costs for both procedures (Tables 7.2 and 7.3) were estimated using the2004 National Collective Contract of Category inclusive of employer’s contributions. We includedthe costs for health operators which the highest number of responding centres had identified asbeing involved in a standard procedure. For DEXA (Table 7.2) the professional figures involved ina standard exam were: the physician (who in the majority of cases observed did not undertakethe measurement but only interviewed the patient), the radiologist (who in the majority of casesobserved undertook the measurement) and administrative staff.

Table 7.2: Mean costs of health personnel for a single DEXA procedure

* Ticket payment: time dedicated by the administerative staff** Medical report: time dedicated by the administerative staff*** Excluded from the cost calculation because present in few cases

Consultation/ticket payment*

Dexa procedure/medical report**

Total

Physicianmean €7,65 €10,90*** €7,65

min-max (€2,19-€10,90) (€5,46-€16,40)

Radiologistmean €3,00 €9,95 €12,95

min-max (€1,53-€7,65) (€5,36-€15,30)

Administrative staffmean €3,04 €2,03 €5,07

min-max (€1,07-€5,07) (€0,68-€3,38)

Totalmean €13,694 €11,93 €25,67

min-max (€4,79-€23,62) (€6,04-€18,68) (€10,83-€42,30)

34

SOURCES OF DATA DATA VALUE

ISTAT, 2008 Italian female population >50 years 12,614,484

ESOPO Study (2004) Prevalence 20%

Clowes et al. 2006

Achilles BUA (T score -0.81)• Sensibility• Specificity• Estimated true positive

96%55%

19.2%

DBM Ad SOS (T score -1.96)• Sensibility• Specificity• Estimated true positive

96%33%

19.2%

For QUS (Table 7.3) the professional figures involved in a standard exam were: the physicianand administrative staff (in 2 out of 10 centres, the cost of the administrative personnel were stillused for the calculation of costs. The lack of response was attributed to difficulty of the intervie-wed to locate data.

Table 7.3: Mean costs of health personnel for a single QUS procedure

* Ticket payment: time dedicated by the administerative staff** Medical report: time dedicated by the administerative staff

The costs of consumed materials for the two procedures were valued as indicated by theresponding centres and when reporting was incomplete we based our calculations on market pri-ces (Table 7.4).

Table 7.4: Costs of consumables for a single DEXA or QUS procedure

*Not applicable

For the costs of the DEXA and QUS equipments (Table 7.5) we chose minimum values (rela-ting to a “basic” technology) and maximum values (representing a “basic technology with addedsoftware”) indicated by the responding centres. The depreciation rate was 8 years considering themaximum use of the technology per year (6.000 procedures per year) as reported by the centres.Depreciation was estimated with the method of straight-line depreciation (straight line deprecia-tion is calculated by taking the purchase or acquisition price of an asset subtracted by the salva-ge value divided by the total expected productive years of the device).

Consumable DEXA cost QUS cost

Toner €0.40 €0.40

Printer cartridge €0.03 €0.03

Paper €0.02 €0.03

Bag €0.01 NA *

Paper roll €0.03 €0.03

Disinfectant €0.01 €0.01

Gloves €0.10 -

Files NA * €0.01

Gel for ultrasound NA * €0.01

Consultation/ticket payment*

QUS procedure/medical report**

Total

Physicianmean €10,90 €9,84 €20,74

min-max (€5,46-€16,40) (€5,46-€10,90)

Administrative staffmean €1,69 €1,69 €3,38

min-max €1,69 €1,69

Totalmean €12,59 €11,53 €24,12

min-max (€7,15-€18,09) (€7,15-€12,59) (€14,30-€30,68)

35

Table. 7.5: Cost of equipment

A single DEXA exam has a direct mean cost (considering personnel, consumables and equip-ment costs) of €30,81 (range: €14,16-49,33€). A single QUS exam (general costs included) has amean direct cost of €26,43 (€16,04-€33,37) (table 7.6).

Table 7.6: Direct mean unit cost for a DEXA and QUS

7.4 Cost effectiveness analysis The Cost effectiveness analysis was carried out to compare a DEXA exam with a QUS exam (Tables7.7a, 7.7b). Our target population for the cost effectiveness analysis is represented by the num-ber of Italian postmenopausal women in the year 2008 (12,614,484 women). According to theISTAT definition34 postmenopausal women are women aged 50 or more.

We estimated direct health costs of the DEXA and QUS procedures from the data of our survey of15 of the 36 centres of sample to which we added the data from the pilot centre in Rome. Costsare reported as means, a minimum cost (basic equipment) and a maximum cost (basic equipmentand additional software/equipment).

DEXA QUS

Personnel €25,67 €24,12

Range (€10,83-€42,30) (€14,30-€30,68)

Equipment €3,07 €0,49

Range (€2,06-€4,08) (€0,42-€0,55)

Consumables €0,60 €0,52

General costs (5%) €1,47 €1,30

Range (€0,67-€2,35) (€0,80-€1,62)

Average total cost €30,81 €26,43

Range (€14,16-€49,33) (€16,04-€33,37)

Equipment Quantity Min Cost Max. Cost Servicing Depreciation

DEXA 1 €58.800 €156.000 €5.000 8

QUS 1 €20.280 €26.400 €0 8

36

Efficacy data, as stated above (par. 7.2), refer to the costs of the osteporosis cases identified byDEXA or QUS in postmenopausal women. We chose to compare efficacy of DEXA at the hip to QUSAchilles and BMD QUS (performed on the phalanges) because both methods are used in ourresponding centres. DEXA at the hip identified 9,8 cases of osteoporosis on a population of 100postmenopausal women. QUS detected 9,4 cases of osteoporosis per 100 postmenopausalwomen3. We subsequently increased the prevalence to 20% for the sensitivity analysis and theestimated true positives were 19,2 cases (table 7.8a,7.8b).

Cost effectiveness analysis shows that DEXA compared with QUS (Achilles BUA, DBM) producesan ICER (cost per each unit of effect) of €1.095 when we consider the mean standard cost. TheICER increases to €3.990 if we consider the higher average cost that includes additional softwarefor DEXA and QUS. DEXA shows to be dominant option when considering the minimum cost (basicequipment). Using the same costs and varying the prevalence from 9.8 to 20 every 100 postme-nopausal women (Table 7.8a,7.8b), DEXA is even more cost effective; ICER’s values dramaticallydecrease (€548 using the average costs of equipment and €1.995 using maximum cost).

Table 7.7a: ICER per identified person with osteoporosis by DEXA compared to QUS – based onthe prevalence reported in the study by Clowes et al (9,8%)

(*) we refer to models Achilles BUA (T score - 0.81) and QUS DBM (T score - 1.96)

Table 7.7b: Sensitivity analysis

(*)we refer to models Achilles BUA (T score - 0.81) and QUS DBM (T score - 1.96)

Table 7.8a: ICER per identified person with osteoporosis by DEXA compared to QUS – based onthe prevalence reported in the ESOPO study (20%)


HIP DEXA(T score - 2.50)

QUS (*) CE

(n. cases identified)ICER

Min cost € 178.621.093 € 202.336.323 -€ 23.715.230 - Dominant

Max cost € 622.272.496 € 420.2945.331 € 201.327.165 - € 3.990

Efficacy 1.236.219 1.185.761 50.458

Technology CostEfficacy

(n. cases identified)C E


HIP DEXA (T score - 2.50)

€ 388.652.252 1.236.219€ 55.251.440 50.458 €1.095

QUS (*) € 333.400.812 1.185.761

37

Technology CostEfficacy

(n. cases identified) CE


HIP DEXA (Tscore - 2.50)

€ 388.652.252 2.522.897€ 55.251.440 100.916 € 548

QUS (*) € 333.400.812 2.421.981

Table 7.8b: Sensitivity analysis


7.5 Results and discussion DEXA is the most used device at outpatient level for the identification of osteoporosis in our

target population and has a higher accuracy when compared with QUS. When using basic DEXAand QUS equipment, our economic evaluation shows that DEXA is the dominant procedure foridentifying osteoporosis cases in a general population of postmenopusal women. As the prevalen-ce of osteoporosis increases, the ICER becomes even more favourable to DEXA. At the same timethe accuracy of QUS increases (there are fewer false positives).

38

HIP DEXA(T score - 2.50)

QUS (*) CE


Min cost € 178.621.093 € 202.336.323 -€ 23.715.230 - Dominant

Max cost € 622.272.496 € 420.945.331 € 201.327.165 - € 1.995

Efficacy 2.522.897 2.421.981 - 100.916 -

39

8. Discussion

To our knowledge our assessment and appraisal, is the first of its kind to be carried out in Italy.There are several novel features of the process, including an albeit incomplete picture of the useof technologies for identification of osteoporosis, regional variations in their use and a survey ofresource use and unit costs involved in operating DEXA and QUS. Our context analysis was basedon a comparatively low response rate of just over 40% but we have no reason to believe that thismay affect our conclusions which are based on evidence from literature and prospective resourcedata gathering.

Our results are relatively straightforward, showing DEXA as a dominant technology and QUSas a poor second best. Although portable, cheaper and possibly safer, the performance of heelQUS is significantly inferior to that of DEXA in a population of asymptomatic postmenopausalwomen. Although this was not the primary aim of our report, QUS is also unsuitable as a scree-ning tool because of its low specificity which would result in a high rate of false positives and thenecessity to repeat the exam with DEXA with high costs.

Interregional variability in the provision of DEXA appeared minimal with the exception of theMarche region and the reasons for this should be investigated. In addition a rolling audit of DEXAreferrals could be carried out to assess need, equity of access and appropriateness of provision.In turn DEXA may not be a contender for mass screening, because of its resource implications andlimited offer. However such a role could be the object of large population trials perhaps compa-ring the performance of DEXA with that of additional risk factors in Italian post menopausalwomen as a means of identifying suitable candidates for extra surveillance or treatment. However,such studies would have to include pharmacological or dietary interventions. The assessment ofthe evidence of the effects of these interventions for osteoporosis were outside the scope of thisreport.

We could not follow the rationale of those authors such as Frost and colleagues (2001) whocalled for a revision of the WHO diagnostic criteria for osteoporosis and osteopenia. Such an actionwould have the certain effect of increasing considerably the number of women identified as osteo-porotic. Otherwise some authors35 suppose that current WHO criteria may have been influencedby occult industry sponsorship of the meetings leading to their promulgation in 1993; this couldhave influenced the change of a simply risk factor (bone loss) in a disease (osteoporosis). In ouropinion classification changes leading to a redefinition of risk factors such as osteoporosis andosteopenia would have to be based on firmer evidence of clinical and economic benefit.

In conclusion we see no reason to recommend changes to the current provision of osteopo-rosis assessment. If further technical advances are made in the development of QUS with conse-quent improvement in performance, these could be taken into consideration in a new report.

41

9. Recommendations

We recommend no change to the current general provision of DEXA services. We suggest thata rolling audit be instituted in DEXA provider centres to assess need, equity and adequacy of pro-vision. Should improvements in the performance of QUS or a promising alternative arise, thisreport should be updated to take these changes into consideration.

10. Funding

Production of this report was made possible by financial contributions from the Italian Ministryof Labour, Health and Social Policies (Ministerial Committee on Medical Devices) and Age.na.s.

Age.na.s. takes sole responsibility for the final form and content of this report. The viewsexpressed herein do not necessarily represent the views of the Italian MLHSP or any regionalgovernment.

43

11. Competing interest declaration

The authors declare that they will not receive either benefits or harms from the publication ofthis report. None of the authors have or have held shares, consultancies or personal relationshipswith any of the producers of the devices assessed in this document.

45

47

Bibliography

1. SBU - Swedish Council on Technology Assessment in Health Care, Summery andconclusion of the report “Osteoporosis – Prevention, identification and treatment.A systematic review”, 2003.

2. Osteporosi: una malattia sociale. Epidemiologia, costi assistenziali, interventi tera-peutici.http://www.ministerosalute.it/resources/static/primopiano/278/scheda%20ISS%20def%20.pdf. (date last accessed 4th December 2008).

3. Clowes JA, Peel NF, Eastell R.Device-specific thresholds to diagnose osteoporosis atthe proximal femur: an approach to interpreting peripheral bone measurements inclinical practice.Osteoporos Int. 2006;17(9):1293-302.

4. WHO Study Group (1994) Assessment of fracture risk and its application to screeningfor postmenopausal osteoporosis. Report of a WHO Study Group. World Health OrganTech Rep Ser 843:1–129.

5. Cooper C, Campion G, Melton LJ 3rd. Hip fractures in the elderly: a world-wide pro-jection. Osteoporos Int. 1992 Nov;2(6):285-9

6. Studio Esopo –Epidemiological Study on the Prevalence of Osteporosis: Crepaldi et al.ASBMER 2004

7. Genant HK, FaulknerKG, Gluer CC, Engelke K, Bone densitometry: current asses-sment. Osteoporosis Int 1993, 51:91-97

8. Lim LS, Hoeksema LJ, Sherin K; ACPM Prevention Practice Committee. Screening forosteoporosis in the adult U.S. population: ACPM position statement on preventivepractice. Am J Prev Med. 2009 Apr;36(4):366-75.

9. Albanese CV, Diessel E, Genant HK. Clinical applications of body composition measu-rements using DXA. J Clin Densitom. 2003 Summer;6(2):75-85.

10. Mauck KF, Clarke BL. Identification , screening, prevention, and treatment of osteo-porosis. Mayo Clin Proc. 2006 May;81(5):662-72.

11. http://www.radiologyinfo.org/en/info.cfm?pg=DEXA (date last accessed 8th June2009)

12. http://courses.washington.edu/bonephys/opbmd.html (date last accessed 8th June2009

13. Dawson-Hughes B, Dallal GE. Effect of radiographic abnormalities on rate of bone lossfrom the spine. Calcif Tissue Int 1990;46:280 –1.

14. Wahner H. Technical aspects and clinical interpretation of bone mineral measure-ments. Public Health Rep. 1989 Sep-Oct;104 Suppl:27-30.

15. Cummings SR, Bates D, Black DM. Clinical Use of Bone Densitometry: ScientificReview JAMA. 2002;288(15):1889-1897

16. http://www.aaos.org/ (date last accessed 29th May 2009)

48

17. Boonen S, Nijs J, Borghs H, Peeters H, Vanderschueren D, Luyten FP. Identifyingpostmenopausal women with osteoporosis by calcaneal ultrasound, metacarpal digi-tal X-ray radiogrammetry and phalangeal radiographic absorptiometry: a comparati-ve study. Osteoporos Int. 2005 Jan;16(1):93-100.

18. Weigert JM. Measuring bone mineral density: let me count the ways! Flat-panel volu-metric CT: another new tool, Acad Radiol. 2009 Apr;16(4):391-3.

19. MacLaughlin E.J., MacLaughlin A.A., Snella K.A., et al. Osteoporosis Screening andEducation in Community Pharmacies using a team approach. Pharmacotherapy 2005;25(3):379-386.

20. Leib ES, Lewiecki EM, Binkley N, Hamdy RC, International Society for ClinicalDensitometry. Official positions of the International Society for Clinical Densitometry.J Clin Densitom. Spring 2004;7:1-6.

21. Mauck KF, Clarke BL.Identification , screening, prevention, and treatment of osteopo-rosis. Mayo Clin Proc. 2006 May;81(5):662-72.

22. Australia and New Zealand Horizon Scanning Network – Quantitative Ultrasound(QUS): Horizon Scanning Report. May 2008. http://www.horizonscanning.gov.au

23. Kraemer DF, Nelson HD, Bauer DC, Helfand M.Economic comparison of diagnosticapproaches for evaluating osteoporosis in older women.Osteoporos Int. 2006Jan;17(1):68-76.

24. Frost, M. L., Blake, G. M. & Fogelman, I. Can the WHO criteria for diagnosing osteo-porosis be applied to calcaneal quantitative ultrasound?Osteoporos Int. 2000 11 (4),321-330.

25. Laabes EP, Vanderjagt DJ, Obadofin MO, Sendeht AJ, Glew RH.Assessment of thebone quality of black male athletes using calcaneal ultrasound: a cross-sectionalstudy. Nutr Metab (Lond). 2008 May 20;5:13.

26. G. Díaz-Guerra, L. Gil-Fraguas, E. Jódar, J. Meneu, E. García, M. Gómez, E. Moreno,F. Hawkins Quantitative Ultrasound of the Calcaneus in Long-Term Liver or CardiacTransplantation Patients. Journal of Clinical Densitometry, Volume 9, Issue 4, Pages469-47.

27. Committee on the Biological Effects of Ionizing Radiations, 1990 Health effects ofexposure to low levels of ionizing radiation: Washington, DC: National AcademyPress.

28. Thomas SR, Kalkwarf HJ, Buckley DD, Heubi JE Effective dose of dual-energy X-rayabsorptiometry scans in children as a function of age.. J Clin Densitom. 2005Winter;8(4):415-22.

29. Frost, M. L., Blake, G. M., Fogelman, I. Quantitative Ultrasound and Bone MineralDensity Are Equally Strongly Associated with Risk Factors for Osteoporosis Journal ofbone and mineral research Volume 16, Number 2, 2001 p.406-416.

30. Dane, C., Dane, B., Cetin, A. and Erginbas, M.(2008) 'The role of quantitative ultra-sound in predicting osteoporosis defined by dual-energy X-ray absorptiometry in pre-and postmenopausal women',Climacteric,11:4,296 - 303.

31. Gemalmaz A., Güzel D., Ceylan C. Diagnostic Performance of QUS for IdentifyingOsteoporosis in Postmenopausal Turkish Women. Turk J Med Sci 2007; 37 (5): 303-309.

32. Schott A.M., Kassaý¨ Koupaý¨ Æ B., Hans D., et al. Should age influence the choiceof quantitative bone assessment technique in elderly women? The EPIDOS study.Osteoporos Int (2004) 15: 196-203.

33. Larijani B., M. Dabbaghmanesh H., Aghakhani S., et al. Correlation of QuantitativeHeel Ultrasonography with central dual-energy X-ray Absorptiometric bone mineraldensity in postmenopausal women. Ultrasound Med 2005 24:941-946.

34. Annuario ISTAT 2008.

35. Napoli M. The marketing of osteoporosis. AJN 2009; 109:58-61.

36. Annuario ISTAT 2007.

37. Cetin A, Ertürk H, Celiker R, Sivri A, Hasçelik Z.The role of quantitative ultrasound inpredicting osteoporosis defined by dual X-ray absorptiometry. Rheumatol Int. 2001Feb;20(2):55-9.

49

51

APPENDIX 1Technologies for bone measurements

Mechanisms of the bone mass lossBones are made of two types of tissue: the cortical (or compact) bone and the trabecular or

cancellous bone (also knows as spongy bone). Cortical bone forms the outer shell of bones whe-reas trabecular bone, that is more metabolically active, forms the inner part. These two types oftissue are differently distributed in the skeleton (e.g. long bones contain more cortical bone whe-reas vertebrae consist mainly of trabecular bone) (SBU 2003).

Bones are constantly formed and resorbed during the normal metabolism. This balance is dueto the activity of osteoclasts (cells that resorb bone tissue) and osteoblasts (cells that form bonetissue) (Australia HS 2008). These cells are regulated by local endocrine factors and mechanicalsolicitations. When the balance is altered towards bone resorption or the formation of new boneis decreased, this causes bone loss and structural damage.

With age the production of bone constructive hormones and the formation of vitamin D in thekidneys both decrease. This is particular important for women as the production of oestrogendecreases with age. These factors, combined with a decrease in physical activity, may lead to aloss of bone tissue in the elderly (SBU 2003).

Bone measurements Bone mass measurements are recognised among the main indicators in the diagnosis and

management of patients with osteoporosis. These measurements are important for monitoringtreatment effects and identifying rapid bone loss (Wahner, 1989). In particular, bone mineral den-sity (BMD expressed as g/cm2) has come to play a major role in both the diagnosis and follow-upof patients with osteoporosis.

In the past decades several different modalities have been developed to measure BMD usingdifferent technological concepts. The firsts applications (1940s) involved plain radiographs obtai-ned by X-rays (e.g., the Singh index that described the trabecular patterns of the bone) (AAOSwebsite). However, osteoporosis generally presents a slow progression and small changes in BMDwere difficult to determine by these techniques. Thus, other technologies with higher diagnosticprecision have been specifically developed to measure BMD. The most common technologies usedin clinical practice, both in the past and at present, are based on the principle of the x-rays tran-smission and in particular, on the measurement of the energy absorbed by calcium that representsthe mineral component of bone. Other technologies recently used are based on sound wave pro-pagation or on the magnetic resonance concepts.

The different sites of measure are often related to the limitation of the particular measure-ment device (e.g. the need to reduce the interference of the soft tissue in the signal propagation).Generally, measures are performed in central or peripheral sites. The femur, the hip, the columnare central sites whereas the peripheral sites are the calcaneus, the radius or the phalanges.

52

TechnologiesDifferent technologies can be used for bone measurements (WHO 2000). The following are

amongst the most common technologies used for assessing bone mass.

Single and Double Photon Absorptiometry (SPA and DPA)SPA was developed in the early 1960s. The site of measurement (radius or calcaneus) is scan-

ned with a single-energy photon beam (generally form 125I) and the BMD is then calculated formthe images. SPA is affected by the amount of soft tissue in the region of interest and can be onlyperformed peripherally since the amount of soft tissue in theses areas is small.

DPA was developed in 1980s and uses a double-energy photon beam (generally from 153Gd).The two energy peaks are absorbed by the soft tissue and by bone respectively. The soft-tissuecomponent is subtracted to determine the BMD. DPA allowed for the first time BMD measurementsin the spine, in the proximal femur and in the whole body.

Both SPA and DPA have limited value in monitoring BMD changes over time since the radio-active source that generates the signal is not constant with time, gradually decays and must bereplaced periodically (Wahner 1987 and AAOS website).

Dual-Energy X-ray Absorptiometry (DEXA or DXA)DEXA is the evolution of DPA and is a well established technology to measure BMD (first

applied about 20 years ago) (Genant 1993). A beam of low-dose x-rays are passed through thearea being measured and are acquired by a detector in the instrument's arm. The system tran-sduces the signal to an image of the skeleton, and analyses the bone density (Cliwes website).Data can be obtained as bone mineral content (BMC) in g/cm or areal bone mineral density (BMD)in g/cm2.

DEXA can be used for lumbar spine (posterior-anterior and lateral projections) and proximalfemur; these are defined central sites or in peripheral sites such as the radius and the calcaneus.Measurements in total body mode are mainly used in order to achieve overall and regional asses-sment of body composition in terms of fat, lean mass and bone (Albanese 2003).

Manufacturers have developed different concepts in order to reduce exposure times and impro-ve image quality. The systems differ also in the methods with which the energy beam is produced:by a pulsating voltage source, or by filtering methods. Also scanning modalities may differ from onedevice to another: some use a simple “pencil beam” to reduce the radiation dose, others use a “fanbeam” that reduces the scan time and improves the image quality.

Quantitative Computed Tomography (QCT)QCT was developed in the late 1970s (Weigert, 2009) as an adaptation of the routine compu-

ted tomography for organ imaging. It uses x-rays to perform a true volumetric measurement ofBMD in mg/cm3 allowing analysis of cortical and trabecular bone separately, so it is less influen-

ced by the changes caused by degenerative disease (which can interfere with DEXA accuracy) (Lim2009). However, QCT is more expensive and time consuming than DEXA and exposes patients toa marked increase in radiation (radiation dose ranges may reach value comparable to chest x-ray,from 25–360 ìSv) (Wahner 1897, Njeha 1999, Cummings 2002).

Quantitative Ultrasound (QUS)QUS uses sound waves to assess some bone features such as microarchitecture (e.g. trabe-

cular architecture) and elasticity. The measurements consist of several parameters such as thespeed of sound and the pattern of absorption of different wavelengths, defined as broadbandultrasound attenuation. Although these parameters, named SOS and BUA, are QUS-specific theycan be used to assess BMD (Mueller 2008).

However QUS can be only used in peripheral sites. This may represent a problem as low bonedensity in central sites may be missed (AAOS website). Moreover, peripheral QUS measurementsare not highly correlated with DEXA measurements acquired in central sites (Mueller 2008).

Radiographic Absorptiometry (RA)RA was developed during the late 1980s as an easy way to determine BMD with plain x-ray.

Two plain radiographies of the hand are taken with slightly different exposure techniques using astandard radiographic unit. The procedure is performed on the non-dominant hand and is compu-ter-assisted. An aluminium wedge is included in each film as a reference. Both films are then digi-talised and analysed by a computer that allows comparisons between the density of the bone andthe density of the reference wedge (Boonen, 2005 and AAOS website).

Magnetic Resonance Imaging (MRI)Interest in MRI for bone measurements has grown in order to overcome the resolution limits

of computed tomography (high resolution CT images can not be obtained without a high radiationdose). MRI does not provide direct information on bone density but can provides informationabout the quality of trabecular bone with some resolution. At present, MRI investigation of theskeleton remains a research procedure because of its high costs and complexity (WHO 2000,Arokoski 2002).

53

Suggested reading

SBU. Osteoporosis – Prevention, Diagnosis and Treatment: A Systematic Review, The SwedishCouncil on Technology Assessment in Health Care, Stockholm, Sweden (2003).

Australia and New Zealand Horizon Scanning Network – Quantitative Ultrasound (QUS): HorizonScanning Report. May 2008. http://www.horizonscanning.gov.au

Wahner H. Technical aspects and clinical interpretation of bone mineral measurements.PublicHealth Rep. 1989 Sep-Oct;104 Suppl:27-30.

http://www.aaos.org/ (date last accessed 29th May 2009)

Genant HK, Faulkner KG, Gluer CC, et al: Bone densitometry: current assessment. Osteoporosis3:S91--S97, 1993

http://www.cliwes.rm.cnr.it/ (date last accessed 29th May 2009)

Albanese CV, Diessel E, Genant HK. Clinical applications of body composition measurementsusing DXA. J Clin Densitom. 2003 Summer;6(2):75-85.

Weigert JM. Measuring bone mineral density: let me count the ways! Flat-panel volumetric CT:another new tool! Acad Radiol. 2009 Apr;16(4):391-3.

Lim LS, Hoeksema LJ, Sherin K; ACPM Prevention Practice Committee. Screening for osteoporo-sis in the adult U.S. population: ACPM position statement on preventive practice. Am J PrevMed. 2009 Apr;36(4):366-75.

Njeha CF, Fuersta T, Hansa D, et al. Radiation exposure in bone mineral density assessmentApplied Radiation and Isotopes Volume 50, Issue 1, January 1999, Pages 215-236

Cummings SR, Bates D, Black DM. Clinical Use of Bone Densitometry: Scientific Review JAMA.2002;288(15):1889-1897

Mueller D, Gandjour A. Cost effectiveness of ultrasound and bone densitometry for osteoporosisscreening in post-menopausal women. Appl Health Econ Health Policy. 2008;6(2-3):113-35.

Boonen S, Nijs J, Borghs H, et al. Identifying postmenopausal women with osteoporosis by cal-caneal ultrasound, metacarpal digital X-ray radiogrammetry and phalangeal radiographic absor-ptiometry: a comparative study. Osteoporos Int. 2005 Jan;16(1):93-100.

54

WHO Scientific Group on the Prevention and Management of Osteoporosis (2000: Geneva,Switzerland) - Prevention and management of osteoporosis: report of a WHO scientific group.(WHO technical report series; 921).

Arokoski MH, Arokoski JP, Vainio P et al. Comparison of DXA and MRI methods for interpretingfemoral neck bone mineral density. J Clin Densitom. 2002 Fall;5(3):289-96.

55

57

Appendix 2

Search strategy by database

CRD Dare:( ( "x ray" OR "X-ray" ) AND ( "dual-energy" OR "dual energy" ) ) AND absorptiometry AND

(osteoporosis OR menopause ) OR postmenopause

( ( DXA OR DEXA ) AND ( osteoporosis OR postmenopause ) ) AND QUS OR "quantitative ultra-sound" AND elderly OR age

PubMed:("1998-01-01"[PDAT] : "2009-01-31"[PDAT]) AND (((ray[All Fields])AND (energy[All Fields])

AND (absorptiometry[All Fields])) OR DXA OR DEXA) AND ((ultrasound[All Fields]) OR (qus[AllFields])) AND (('osteoporosis'[MeSH Terms] OR 'osteoporosis'[All Fields]) OR ('menopause'[MeSHTerms] OR 'menopause'[All Fields]) OR ("postmenopause"[MeSH Terms] OR 'postmenopause'[AllFields]))) AND(humans[MeSH Terms] AND (English[lang] OR Italian[lang]))

EMBASE:(((('x ray'/exp OR 'x ray') OR ('x ray'/exp OR 'xray')) OR (('x ray'/exp OR 'x ray') OR ('x ray'/exp

OR 'x ray'))) OR ((('x-ray'/exp OR 'x-ray') OR ('x-ray'/exp OR 'x-ray')) OR (('x-ray'/exp OR 'x-ray')OR ('x-ray'/exp OR 'x-ray')))) AND ('dual-energy' OR 'dual energy') AND ((('absorptiometry'/expOR'absorptiometry') OR ('absorptiometry'/exp OR 'absorptiometry')) OR (('absorptiometry'/exp OR'absorptiometry') OR ('absorptiometry'/exp OR 'absorptiometry'))) AND (((('osteoporosis'/exp OR'osteoporosis') OR ('osteoporosis'/exp OR 'osteoporosis') ) OR (('osteoporosis'/exp OR 'osteoporo-sis') OR ('osteoporosis'/exp OR 'osteoporosis'))) AND (('primary menopause') OR(('menopause'/exp OR 'menopause') OR ('menopause'/exp OR 'menopause')))) AND [1998-2008]/py

((('absorptiometry'/exp OR 'absorptiometry') AND (x AND ray OR 'x -ray')) OR ((x AND ray) OR(('x-ry'/exp OR 'x-ray')) AND ('absorptiometry'/exp OR 'absorptiometry'))) AND ('densitometry'/expOR 'densitometry') AND ('menopause'/exp OR 'menopause') AND [english]/lim AND [humans]/limAND [1998-2008]/ py AND [1998-2008]/py

('quantitative ultrasound' OR qus) AND ('osteoporosis'/exp OR 'osteoporosis') AND ('menopau-se'/exp OR 'menopause') AND [english]/lim AND [humans]/limAND [1998-2008]/py

58

Cochrane Library(("x ray" or "X-ray") and ("dual-energy" or "dual energy")) and absorptiometry and (osteopo-

rosis or postmenopause) and (DXA or DEXA) and (osteoporosis or postmenopause) and QUS and(osteoporosis or postmenopause) or "quantitative ultrasound"

Years: 1998-2009.

59

Appendix 3

QUADAS tool

Quality Assessment of Diagnostic Accuracy StudiesWhiting et al. http://www.biomedcentral.com/1471-2288/3/25

61

Appendix 4

List of excluded studies with reasons for exclusion

Studies excluded because they were not reporting original data Barrett-Connor E. Black D. Bonjour J.-P. et al. Prevention and management of osteoporosis. WorldHealth Organization - Technical Report Series (2003) -:921 (i-192).

Pluskiewicz W., Drozdzowska B., Weiss M., Discrimination of proximal hip fracture by quantitativeultrasound measurement at the radius [1] (multiple letters), Osteoporosis International, (2002)13:3 (265).

Studies excluded because not carried out on humansNo study identified

Studies excluded because there was no direct comparison betweenDEXA and QUSAgency for Healthcare Research and Quality, Osteoporosis in postmenopausal women: identificationand monitoring, Rockville, MD: Agency for Healthcare Research and Quality (AHRQ), EvidenceReport/Technology Assessment 28, Agency for Healthcare Research and Quality (AHRQ), 2001

Alenfeld F.E. Engelke K. Schmidt D. et al. Diagnostic agreement of two calcaneal ultrasound devi-ces: The Sahara bone sonometer and the Achilles. British Journal of Radiology (2002) 75:899(895-902).

Ben Sedrine W, Broers P, Devogelaer J P, et al, Interest of a prescreening questionnaire to reducethe cost of bone densitometry, Osteoporosis International, 2002, 13(5):434-442

Blue Cross Blue Shield Association, Screening for vertebral fracture with dual x-ray absorptiome-try, Chicago IL: Blue Cross Blue Shield Association (BCBS), TEC Assessment 20(14), Blue CrossBlue Shield Association (BCBS), 2006, p.20

Blue Cross Blue Shield Association, Vertebral assessment using dual-energy x-ray absorptiometryfor osteoporotic fracture risk assessment, Chicago IL: Blue Cross Blue Shield Association (BCBS)XSE: TEC Assessment 19(14), Blue Cross Blue Shield Association (BCBS), 2004, p 19

Braga Jr. J.W.R. Neves R.M.S. et al. Pinheiro M.M., Prevalence of low trauma fractures in long-termkidney transplant patients with preserved renal function Brazilian Journal of Medical and BiologicalResearch (2006) 39:1 (137-147).

62

Brooke-Wavell K, Jones PR, Hardman AE. et al, Commencing, continuing and stopping brisk wal-king: effects on bone mineral density, quantitative ultrasound of bone and markers of bone meta-bolism in postmenopausal women., Osteoporosis international : a journal established as result ofcooperation between the European Foundation for Osteoporosis and the National OsteoporosisFoundation of the USA, 2001, 12(7):581-7, ID: CN-00424858

Caraballo P J, Gabriel S E, Castro M R. et al, Changes in bone density after exposure to oral anti-coagulants: a meta-analysis, Osteoporosis International, 1999, 9(5): 441-448

Cepollaro C. Gonnelli S. Montagnani A. et al. In vivo performance evaluation of the Achilles insightQUS device. Journal of Clinical Densitometry (2005) 8:3 (341-346).

Chen J.S. Seibel M.J. Zochling J. et al. Calcaneal ultrasound but not bone turnover predicts frac-tures in vitamin D deficient frail elderly at high risk of falls. Calcified Tissue International (2006)79:1 (37-42).

Damilakis J., Stratakis J., Perisinakis K. et al, Broadband ultrasound attenuation imaging: Algorithmdevelopment and clinical assessment of a region growing technique, Physics in Medicine andBiology, (2002) 47:2 (315-325).

Devine A. Dick I. Islam A.F.M. et al. Protein consumption is an important predictor of lower limbbone mass in elderly women. Am J Clin Nutr (2005) 81:6 (1423-1428).

Doren M, Nilsson J A, Johnell O, Effects of specific post-menopausal hormone therapies on bonemineral density in post-menopausal women: a meta-analysis, Human Reproduction, 2003,18(8):1737-1746

Drozdzowska B., Pluskiewicz W., The ability of quantitative ultrasound at the calcaneus to identifypostmenopausal women with different types of nontraumatic fractures, Ultrasound in Medicine andBiology, (2002) 28:11-12 (1491-1497).

Feltrin G.P., Nardin M., Marangon A. et al, Quantitative ultrasound at the hand phalanges:Comparison with quantitative computed tomography of the lumbar spine in postmenopausalwomen, European Radiology, (2000) 10:5 (826-831).

Frost M.L., Blake G.M, Fogelman I. Can the WHO criteria for diagnosing osteoporosis be appliedto calcaneal quantitative ultrasound?, Osteoporosis International (2000) 11:4 (321-330).

Giguere Y., Dodin S., Blanchet C. et al, The association between heel ultrasound and hormonereplacement therapy is modulated by a two-locus vitamin D and estrogen receptor genotype,Journal of Bone and Mineral Research, (2000) 15:6 (1076-1084).

Giroux S. Elfassihi Cardinal G. L. LRP5 coding polymorphisms influence the variation of peak bonemass in a normal population of French-Canadian women Bone (2007) 40:5 (1299-1307).

Glynn A.W., Michaelsson K., Lind P.M. et al, Organochlorines and bone mineral density in Swedishmen from the general population, Osteoporosis International, (2000) 11:12 (1036-1042).

Goeree R, Blackhouse G, Adachi J, Cost-effectiveness of alternative treatments for women withosteoporosis in Canada, Current Medical Research and Opinion, 2006, 22(7):1425-1436

Gonnelli S. Martini G. Caffarelli C. et al. Teriparatide's effects on quantitative ultrasound parame-ters and bone density in women with established osteoporosis. Osteoporosis International (2006)17:10 (1524-1531).

Grabe D.W. Cerulli J. Stroup J.S. et al Comparison of the Achilles express ultrasonometer with cen-tral dual-energy X-ray absorptiometry. Annals of Pharmacotherapy (2006) 40:5 (830-835).

Hadji P., Hars O., Gorke K., Quantitative ultrasound of os calcis in postmenopausal women withspine and hip fractures, Journal of Clinical Densitometry, (2000) 3:3 (233-239).

Hagenfeldt K, Johansson C, Johnell O, Ljunggren, et al, Osteoporosis - prevention, identificationand treatment, Stockholm: Swedish Council on Technology Assessment in Health Care (SBU),Report no. 165, Swedish Council on Technology Assessment in Health Care (SBU), 2003.

Hans D., Allaoua S., Genton L., et al, Is time since hip fracture influencing the discrimination bet-ween fractured and nonfractured subjects as assessed at the calcaneum by three technologicallydifferent quantitative ultrasound devices?, Calcified Tissue International (2002) 71:6 (485-492).

Harrison E J, Adams J E , Application of a triage approach to peripheral bone densitometry redu-ces the requirement for central DXA but is not cost effective, Calcified TIssue International, 2006,79(4): 199-206

Jaffe J.S., Timell A.M., Gulanski B.I., Prevalence of low bone density in women with developmen-tal disabilities, Journal of Clinical Densitometry, (2001) 4:1 (25-29).

Jin H., Lu Y., Chen M.-H., Reduction of sampling bias of odds ratios for vertebral fractures usingpropensity scores. Osteoporosis International (2006) 17:4 (507-520).

Jones G, Riley M, Couper D, Dwyer , Water fluoridation, bone mass and fracture: a quantitative over-view of the literature, Australian and New Zealand Journal of Public Health, 1999, 23(1): 34-40

Karpf D B, Shapiro D R, Seeman E, Ensrud K E, et al, Prevention of nonvertebral fractures by alen-dronate: a meta-analysis, JAMA, 1997, 277(14) : 1159-1164

Kelley G A, Aerobic exercise and bone density at the hip in postmenopausal women: a meta-ana-lysis, Preventive Medicine, 1998, 27(6): 798-807

Kelley G A, Exercise and regional bone mineral density in postmenopausal women: a meta-analy-tic review of randomized trials, American Journal of Physical Medicine and Rehabilitation, 1998,7(1): 76-87

Kelley G A, Kelley K S, Tran Z V, Resistance training and bone mineral density in women: a meta-analysis of controlled trials, American Journal of Physical Medicine and Rehabilitation, 2001,80(1):65-77

63

Kelley G, Aerobic exercise and lumbar spine bone mineral density in postmenopausal women: ameta-analysis, Journal of the American Geriatrics Society, 1998, 46(2):143-152

Kim C.-H., Kim Y.I., Choi C.S., et al, Prevalence and risk factors of low quantitative ultrasoundvalues of calcaneus in Korean elderly women, Ultrasound in Medicine and Biology, (2000) 26:1(35-40).

Kim H.-S. Kim J.-S. Kim N.S.Association of vitamin D receptor polymorphism with calcaneal broad-band ultrasound attenuation in Korean postmenopausal women with low calcium intake BritishJournal of Nutrition (2007) 98:5 (878-881).

Martini G. Valenti R. Gennari L. et al. Dual X-ray and laser absorptiometry of the calcaneus:Comparison with quantitative ultrasound and dual-energy X-ray absorptiometry. Journal of ClinicalDensitometry (2004) 7:3 (349-354).

Mauloni M., Ventura V., De Aloysio D., et al, Multicenter Italian study on the bone mass ultrasono-graphic evaluation at the phalanges in the climacteric (vumof-clife), Italian Journal of Mineral andElectrolyte Metabolism (2000) 14:2-4 (23-26).

Mobley L R, Hoerger T J, Wittenborn J S, Galuska D A , Cost-effectiveness of osteoporosis scree-ning and treatment with hormone replacement therapy raloxifene, or alendronate, MedicalDecision Making, 2006, 26: 194-206.

Modelska K, Cummings S, Tibolone for postmenopausal women: systematic review of randomizedtrials, Journal of Clinical Endocrinology and Metabolism, 2002, 87(1):16-23

National Institute for Clinical Excellence, “Bisphosphonates (alendronate, etidronate, risedronate),selective oestrogen receptor modulators (raloxifene) and parathyroid hormone (teriparatide) forthe secondary prevention of osteoporotic fragility fractures in postmenopausal women”,Technology Appraisal Guidance, No.87, National Institute for Clinical Excellence (NICE), 2005, p.51

Nelson H D, Helfand M., Screening for postmenopausal osteoporosis, Rockville, MD: Agency forHealthcare Research and Quality (AHRQ), Preventive Services Task Force Systematic EvidenceReview No. 17, Agency for Healthcare Research and Quality (AHRQ), 2002, p 84

Otten B.J., Smals A.G.H., Van den Bergh J.P.W., Measuring skeletal changes with calcaneal ultra-sound imaging in healthy children and adults: The influence of size and location of the region ofinterest, Osteoporosis International (2001) 12:11 (970-979).

Pfister A K, Welch C A, Lester M D , et al, Cost-effectiveness strategies to treat osteoporosis inelderly women, Southern Medical Journal, 2006, 99(2):123-131.

Pluskiewicz W., Drozdzowska B., Quantitative ultrasound (qus) at the calcaneus and hand phalan-ges in polish healthy postmenopausal women, Ultrasound in Medicine and Biology, (2001) 27:3(373-377).

64

Pongchaiyakul C, Nguyen N D, Pongchaiyakul C, Nguyen T V, Development and validation of a newclinical risk index for prediction of osteoporosis in Thai women, Journal of the Medical Associationof Thailand, 2004, 87(8):910-916

Ringertz H, Marshall D, Johansson C, Johnell O, et al, Bone density measurement: a systematicreview, Journal of Internal Medicine, 1997, 241(Supplement 739): 1-60

Rud B, Hilden J, Hyldstrup L, Hrobjartsson A, Performance of the Osteoporosis Self-AssessmentTool in ruling out low bone mineral density in postmenopausal women: a systematic review,Osteoporosis International, 2007, 18(9): 1177-1187

Schott A M, Ganne C, Hans D, Monnier G, et al, Which screening strategy using BMD measure-ments would be most cost effective for hip fracture prevention in elderly women? A decision ana-lysis based on a Markov model, Osteoporosis International, 2007, 18(2):143-151

Schwenkglenks M, Lippuner K , Simulation-based cost-utility analysis of population screening-based alendronate use in Switzerland, Osteoporosis International, 2007, 18(11): 1481-1491

Silberstein E.B. Levin L.L. Fernandez-Ulloa M. Bone Mineral Density (BMD) Assessment of CentralSkeletal Sites From Peripheral BMD and Ultrasonographic Measurements: An Improved SolutionEmploying Age and Weight in Type 3 Journal of Clinical Densitometry (2006) 9:3 (323-328).

Solomon D H, Kuntz K M , Should postmenopausal women with rheumatoid arthritis who are star-ting corticosteroid treatment be screened for osteoporosis? A cost-effectiveness analysis, Arthritisand Rheumatism, 2000, 43(9):1967-1975

Solomon D.H.Will absolute fracture risk prediction facilitate treatment of osteoporosis? NatureClinical Practice Endocrinology and Metabolism (2008) 4:9 (480-481).

Sosa M., Saavedra P., Gomez-Alonso C., et al Postmenopausal women with Colles' fracture havebone mineral density values similar to those of controls when measured with calcaneus quantita-tive ultrasound. European Journal of Internal Medicine (2005) 16:8 (561-566).

Stiger F., Brandstrom H., Gillberg P.et al. Association between repeat length of exon 1 CAG micro-satellite in the androgen receptor and bone density in men is modulated by sex hormone levels A.International (2008) 82:6 (427-435).

Tao B. Liu J.-M. Li X.-Y.An assessment of the use of quantitative ultrasound and the OsteoporosisSelf-Assessment Tool for Asians in determining the risk of nonvertebral fracture in postmenopau-sal Chinese women. Journal of Bone and Mineral Metabolism (2008) 26:1 (60-65).

Tashiro A. Kakuta H. Tanaka N. Relationship between health-related quality of life domains andbone status in postmenopausal Japanese women. Menopause (2006) 13:5 (846-849).

65

Van Der Voort D J, Brandon S, Dinant G J, van Wersch J W, Screening for osteoporosis using easi-ly obtainable biometrical data: diagnostic accuracy of measured, self-reported and recalled BMI,and related costs of bone mineral density measurements, Osteoporosis International, 2000, 11(3):233-239

Visentin P, Ciravegna R, Fabris F, Estimating the cost per avoided hip fracture by osteoporosis tre-atment in Italy, Maturitas, 1997, 26(3): 185-192

Wolff I, van Croonenborg J J, Kemper H C, et al, The effect of exercise training programs on bonemass: a meta-analysis of published controlled trials in pre- and postmenopausal women,Osteoporosis International, 1999, 9(1):1-12.

Zajickova K., Zemanova A., Hill M. et al. Is A163G polymorphism in the osteoprotegerin geneassociated with heel velocity of sound in postmenopausal women? Physiol. Res. (2008) 57:SUPPL.1 (S153-S157).

Zehnacker C H, Bemis-Dougherty A, Effect of weighted exercises on bone mineral density in postmenopausal women: a systematic review, Journal of Geriatric Physical Therapy, 2007, 30(2): 79-88

Studies excluded because they were not carried out on primary osteo-porosis, postmenopausal caucasian women and did not report specifici-ty and sensitivity of DEXA and QUS Abbott T A, Mucha L, Manfredonia D, Schwartz E N, et al, Efficient patient identification strate-gies for women with osteoporosis, Journal of Clinical Densitometry, 1999, 2(3):223-230

Adler R.A., Funkhouser H.L., Holt C.M., Utility of heel ultrasound bone density in men, Journal ofClinical Densitometry, (2001) 4:3 (225-230).

Albanese CV, Cepollaro C, de Terlizzi F, Brandi ML, Passariello R. Performance of five phalangealQUS parameters in the evaluation of gonadal-status, age and vertebral fracture risk compared withDXA. Ultrasound Med Biol 2009; 35(4):537-44.

Alenfeld F.E., Diessel E., Brezger M., et al, Detailed analyses of periarticular osteoporosis in rheu-matoid arthritis, Osteoporosis International, (2000) 11:5 (400-407).

Anonymous, Precision of quantitative ultrasound: comparison of three commercial scanners.,Bone, 2000, 27(1):139-143, ID: CN-00424074

Appetecchia M. Effects on bone mineral density by treatment of benign nodular goiter with mildlysuppressive doses of L-thyroxine in a cohort women study. Horm Res 2005; 64(6):293-8.

Arici M., Erturk H., Altun B., et al, Bone mineral density in haemodialysis patients: A comparativestudy of dual-energy X-ray absorptiometry and quantitative ultrasound, Nephrology DialysisTransplantation, (2000) 15:11 (1847-1851).

66

Assantachai P. Angkamat W. Pongpim P. et al Risk factors of osteoporosis in institutionalized olderThai people. Osteoporosis International (2006) 17:7 (1096-1102).

Assantachai P. Sriussadaporn S. Thamlikitkul V. Body composition: Gender-specific risk factor ofreduced quantitative ultrasound measures in older people. Osteoporosis International (2006) 17:8(1174-1181).

Baca EA, Ulibarri VA, Scariano JK et al. Increased serum levels of N-telopeptides (NTx) of bonecollagen in postmenopausal Nigerian women. Calcif Tissue Int 1999; 65(2):125-8.

Baron YM, Brincat M, Galea R. Increased reduction in bone density and skin thickness in postme-nopausal women on long-term corticosteroid therapy. A suggested role for estrogen add back the-rapy. Adv Exp Med Biol 1999; 455:429-36.

Baron YM, Brincat MP, Galea R. Increased reduction in bone density and skin thickness in postme-nopausal women taking long-term corticosteroid therapy: a suggested role for estrogen add-backtherapy. Climacteric 1999; 2(3):189-96.

Bemben DA, Langdon DB. Relationship between estrogen use and musculoskeletal function inpostmenopausal women. Maturitas 2002; 42(2):119-27.

Blanckaert F, Cortet B, Coquerelle P, Flipo RM, Duquesnoy B, Delcambre B. Ultrasound velocitythrough the phalanges in normal and osteoporotic patients. Calcif Tissue Int 1999; 64(1):28-33.

Bolanowski M. Pluskiewicz W. Skrzek A. Beneficial effects of Tai Chi on women's skeletal statusassessed by quantitative ultrasound at the hand phalanges One-year follow-up study. Advancesin Clinical and Experimental Medicine (2007) 16:5 (675-681).

Bolanowski M. Pluskiewicz W. Skrzek A. et al, Bone properties assessed by quantitative ultrasoundat the hand phalanges in women exercising Tai Chi. Human Movement (2006) 7:2 (162-167).

Brukx LJ, Waelkens JJ, Evaluation of the usefulness of a quantitative ultrasound device in scree-ning of bone mineral density in children., Annals of human biology, 2003, 30(3): 304-15 , ID: CN-00439096

Camozzi V. Carraro V. Zangari M. et al. Use of quantitative ultrasound of the hand phalanges inthe identification of two different osteoporotic syndromes: Cushing's syndrome and postmeno-pausal osteoporosis. Journal of Endocrinological Investigation (2004) 27:6 (510-515).

Camozzi V. Lumachi F. Mantero F. et al. Phalangeal quantitative ultrasound technology and dualenergy X-ray densitometry in patients with primary hyperparathyroidism: Influence of sex andmenopausal status. Osteoporosis International (2003) 14:7 (602-608).

Chappard C. Roux C. Laugier P. et al. Bone status in primary hyperparathyroidism assessed byregional bone mineral density from the whole body scan and QUS imaging at calcaneus. JointBone Spine (2006) 73:1 (86-94).

67

Daly R.M. Bass S.L. Lifetime sport and leisure activity participation is associated with greater bonesize, quality and strength in older men Osteoporosis International (2006) 17:8 (1258-1267).

Damilakis J, Papadokostakis G, Perisinakis K, Maris TG, Karantanas AH. Hip fracture discriminationby the Achilles Insight QUS imaging device. Eur J Radiol 2007; 63(1):59-62.

Damilakis J, Perisinakis K, Kontakis G, Vagios E, Gourtsoyiannis N. Effect of lifetime occupationalphysical activity on indices of bone mineral status in healthy postmenopausal women. Calcif TissueInt 1999; 64(2):112-6.

Damilakis J., Papadokostakis G., Perisinakis K., et al, Discrimination of hip fractures by quantitati-ve ultrasound of the phalanges and the calcaneus and dual X-ray absorptiometry, EuropeanJournal of Radiology, (2004) 50:3 (268-272).

Damilakis J., Papadokostakis G., Vrahoriti H. , et al, Ultrasound velocity through the cortex of pha-langes, radius, and tibia in normal and osteoporotic postmenopausal women using a new multisi-te quantitative ultrasound device, Investigative Radiology, (2003) 38:4 (207-211).

Davutoglu V, Yilmaz M, Soydinc S et al. Mitral annular calcification is associated with osteoporosisin women. Am Heart J 2004; 147(6):1113-6.

Dick I.M. Devine A. Marangou A. et al. Apolipoprotein E4 is associated with reduced calcanealquantitative ultrasound measurements and bone mineral density in elderly women. Bone (2002)31:4 (497-502).

Dos Santos RE, Aldrighi JM, Lanz JR, Ferezin PC, Marone MM. Relationship of body fat distribu-tion by waist circumference, dual-energy X-ray absorptiometry and ultrasonography to insulin resi-stance by homeostasis model assessment and lipid profile in obese and non-obese postmenopau-sal women. Gynecol Endocrinol 2005; 21(5):295-

Douchi T, Ijuin M, Ijuin T, Ijuin Y. Relationship of ultrasonographic endometrial thickness and ute-rine size to bone mineral density in postmenopausal women. Maturitas 2004; 48(3):219-23.

Durosier C., Hans D., Krieg M.A., et al, Defining Risk Thresholds for a 10-Year Probability of HipFracture Model That Combines Clinical Risk Factors and Quantitative Ultrasound: Results Using theEPISEM Cohort, Journal of Clinical Densitometry, (2008) 11:3 (397-403).

Ekman A., Michaelsson K., Petren-Mallmin M., et al, DXA of the hip and heel ultrasound but notdensitometry of the fingers can discriminate female hip fracture patients from controls: A compa-rison between four different methods, Osteoporosis International, (2001) 12:3 (185-191).

Engelke K, Kemmler W, Lauber D, Beeskow C, Pintag R, Kalender WA. Exercise maintains bonedensity at spine and hip EFOPS: a 3-year longitudinal study in early postmenopausal women.Osteoporos Int 2006; 17(1):133-42.

Falchetti A, Sferrazza C, Cepollaro C et al. FokI polymorphism of the vitamin D receptor gene cor-relates with parameters of bone mass and turnover in a female population of the Italian island ofLampedusa. Calcif Tissue Int 2007; 80(1):15-20..

68

Frediani B, Acciai C, Falsetti P et al. Calcaneus ultrasonometry and dual-energy X-ray absorptio-metry for the evaluation of vertebral fracture risk. Calcif Tissue Int 2006; 79(4):223-9.

Frediani B. Falsetti P. Baldi F. et al. Effects of 4-year treatment with once-weekly clodronate onprevention of corticosteroid-induced bone loss and fractures in patients with arthritis: Evaluationwith dual-energy X-ray absorptiometry and quantitative ultrasound. Bone (2003) 33:4 (575-581).

Frost M.L., Blake G.M., Fogelman I., A comparison of fracture discrimination using calcaneal quan-titative ultrasound and dual X-ray absorptiometry in women with a history of fracture at sites otherthan the spine and hip, Calcified Tissue International, (2002) 71:3 (207-211)

Frost M.L., Blake G.M., Fogelman I., Does quantitative ultrasound imaging enhance precision anddiscrimination?, Osteoporosis International, (2000) 11:5 (425-433).

Frost ML, Blake GM, Fogelman I., Does quantitative ultrasound imaging enhance precision anddiscrimination?, Osteoporosis international : a journal established as result of cooperation betwe-en the European Foundation for Osteoporosis and the National Osteoporosis Foundation of theUSA, 2000, 11 (5): 425-33, ID: CN-00298465

Gluer M.G. Minne H.W. Gluer C.-C., Prospective identification of postmenopausal osteoporoticwomen at high vertebral fracture risk by radiography, bone densitometry, quantitative ultrasound,and laboratory findings: Results from the PIOS study. Journal of Clinical Densitometry (2005) 8:4(386-395).

Gnudi S., Ripamonti C., Quantitative ultrasound at the phalanxes discriminates osteoporoticwomen with vertebral but not with hip fracture, Ultrasound in Medicine and Biology, (2004) 30:3(357-361).

Goemaere S. Zmierczak H. Van Pottelbergh I. et al. Ability of peripheral bone assessments to pre-dict areal bone mineral density at hip in community-dwelling elderly men. Journal of ClinicalDensitometry (2002) 5:3 (219-228).

Gómez Acotto C, Schott AM, Hans D, et al, Hyperthyroidism influences ultrasound bone measure-ment on the Os calcis., Osteoporosis international : a journal established as result of cooperationbetween the European Foundation for Osteoporosis and the National Osteoporosis Foundation ofthe USA, 1998, 8(5): 455-9, ID: CN-00157878

Gonnelli S. Cepollaro C. Gennari L. et al.. Quantitative ultrasound and dual-energy X-ray absor-ptiometry in the prediction of fragility fracture in men. Osteoporosis International (2005) 16:8(963-968).

Greenfield D.M., Eastell R., Risk factors for ankle fracture, Osteoporosis International, (2001) 12:2(97-103).

Gudmundsdottir S.L., Indridason O.S., Franzson L., et al, Age-related decline in bone mass mea-sured by dual-energy X-ray absorptiometry and quantitative ultrasound in a population-basedsample of both sexes: Identification of useful ultrasound thresholds for osteoporosis screening,Journal of Clinical Densitometry, (2005) 8:1 (80-86).

69

Guglielmi G. De Terlizzi F. Aucella F. et al Quantitative ultrasound technique at the phalanges indiscriminating between uremic and osteoporotic patients. European Journal of Radiology (2006)60:1 (108-114).

Hailey D, Sampietro-Colom L, Marshall D, Rico R, et al, The effectiveness of bone density measu-rement and associated treatments for prevention of fractures: an international collaborativereview, International Journal of Technology Assessment in Health Care, 1998, 14(2):237-254

Hayman S.R., Drake W.M., Kendler D.L., et al, North American male reference population for speedof sound in bone at multiple skeletal sites, Journal of Clinical Densitometry, (2002) 5:1 (63-71).

Hien V.T.T. Khan N.C. Lam N.T. et al. Determining the prevalence of osteoporosis and related fac-tors using quantitative ultrasound in Vietnamese adult women. American Journal of Epidemiology(2005) 161:9 (824-830).

Ikeda Y., Iki M., Morita A., Age-specific values and cutoff levels for the identification of osteopo-rosis in quantitative ultrasound measurements at the calcaneus with SAHARA in healthy Japanesewomen: Japanese Population-Based Osteoporosis (JPOS) Study, Calcified Tissue International(2002) 71:1 (1-9).

Imashuku Y. Takada M. Murata K. Comparisons of bone mass measurements on various skeletalsites including quantitative ultrasonography of the calcaneus for assessing age-related losses,their correlations, and diagnostic agreement using the Japanese and WHO criteria for osteoporo-sis Radiation Medicine - Medical Imaging and Radiation Oncology (2007) 25:4 (148-154).

Ingle B.M. Machado A.B.C. Pereda C.A. et al Monitoring alendronate and estradiol therapy withquantitative ultrasound and bone mineral density. Journal of Clinical Densitometry (2005) 8:3(278-286).

Ingle B.M., Thomas W.E.G., Eastell R., Differential effects of primary hyperparathyroidism on ultra-sound properties of bone, Osteoporosis International (2002) 13:7 (572-578).

Jansen T.L.Th.A. Aarts M.H.M. Zanen S. et al. Risk assessment for osteoporosis by quantitativeultrasound of the heel in ankylosing spondylitis. Clinical and Experimental Rheumatology (2003)21:5 (599-604).

Jorgensen HL, Kusk P, Madsen B, Fenger M, Lauritzen JB. Serum osteoprotegerin (OPG) and theA163G polymorphism in the OPG promoter region are related to peripheral measures of bone massand fracture odds ratios. J Bone Miner Metab 2004; 22(2):132-8.

Kang C, Speller R. Comparison of ultrasound and dual energy X-ray absorptiometry measurementsin the calcaneus. Br J Radiol 1998; 71(848):861-7

Karavitaki N, Ioannidis G, Giannakopoulos F, Mavrokefalos P, Thalassinos N. Evaluation of bonemineral density of the peripheral skeleton in pre- and postmenopausal women with newly diagno-sed endogenous Cushing's syndrome. Clin Endocrinol (Oxf) 2004; 60(2):264-70.

70

Kelley G A, Kelley K S, Tran Z V , Exercise and bone mineral density in men: a meta-analysis,Journal of Applied Physiology, 2000, 88(5):1730-1736

Kemmler W, Engelke K, Weineck J, Hensen J, Kalender WA. The Erlangen Fitness OsteoporosisPrevention Study: a controlled exercise trial in early postmenopausal women with low bone den-sity-first-year results. Arch Phys Med Rehabil 2003; 84(5):673-82.

Knapp KM, Blake GM, Fogelman I, Doyle DV, Spector TD. Multisite quantitative ultrasound: Colles'fracture discrimination in postmenopausal women. Osteoporos Int 2002; 13(6):474-9.

Knapp KM, Blake GM, Spector TD, Fogelman I. Multisite quantitative ultrasound: precision, age-and menopause-related changes, fracture discrimination, and T-score equivalence with dual-ener-gy X-ray absorptiometry. Osteoporos Int 2001; 12(6):456-64.

Ladinsky GA, Vasilic B, Popescu AM et al. Trabecular structure quantified with the MRI-based vir-tual bone biopsy in postmenopausal women contributes to vertebral deformity burden indepen-dent of areal vertebral BMD. J Bone Miner Res 2008; 23(1):64-74.

Langton C M, Langton D K, Beardsworth S A , Comparison of accuracy and cost effectiveness ofclinical criteria and BUA for referral for BMD assessment by DXA in osteoporotic and osteopenicperimenopausal subjects, Technology and Health Care, 1999, 7(5):319-330

Law P.P.W, Calcaneal quantitative ultrasound measurements: A cross sectional study among elder-ly Chinese in a care and attention home complex and in the community, Hong Kong PhysiotherapyJournal, (2002) 20:- (2-5).

Lopez-Rodriguez F. Mezquita-Raya P. De Dios Luna J. et al. Performance of quantitative ultrasoundin the discrimination of prevalent osteoporotic fractures in a bone metabolic unit Bone (2003) 32:5(571-578).

Manassiev NA, Godsland IF, Crook D, Proudler AJ, Whitehead MI, Stevenson JC. Effect of postme-nopausal oestradiol and dydrogesterone therapy on lipoproteins and insulin sensitivity, secretionand elimination in hysterectomised women. Maturitas 2002; 42(3):233-42.

Marin F, Lopez-Bastida J, Diez-Perez A, Sacristan JA. Bone mineral density referral for dual-ener-gy X-ray absorptiometry using quantitative ultrasound as a prescreening tool in postmenopausalwomen from the general population: a cost-effectiveness analysis. Calcif Tissue Int 2004;74(3):277-83.

Martin JC, Campbell MK, Reid DM. A comparison of radial peripheral quantitative computed tomo-graphy, calcaneal ultrasound, and axial dual energy X-ray absorptiometry measurements in womenaged 45-55 yr. J Clin Densitom 1999; 2(3):265-73.

Meszaros S, Ferencz V, Csupor E et al. Comparison of the femoral neck bone density, quantitativeultrasound and bone density of the heel between dominant and non-dominant side. Eur J Radiol2006; 60(2):293-8.

71

Miller PD, Siris ES, Barrett-Connor E et al. Prediction of fracture risk in postmenopausal whitewomen with peripheral bone densitometry: evidence from the National Osteoporosis RiskAssessment. J Bone Miner Res 2002; 17(12):2222-30.

Montagnani A., Gonnelli S., Cepollaro C., et al, Usefulness of bone quantitative ultrasound in mana-gement of osteoporosis in men, Journal of Clinical Densitometry, (2001) 4:3 (231-237).

Moschinos G, Yannis M, Skeletal site-depent response of bone mineral density and quantitativeultrasound parameters following a 12-month dietary intervention using dairy products fortifiedwith calcium and vitamin D: the Postmenopausal Health Study, British Journal of Nutrition 2006;96, 1140-1148

Naganathan V., Zochling J., March L., et al, Peak bone mass is increased in the hip in daughtersof women with Osteoarthritis, Bone, (2002) 30:1 (287-292).

Nagata-Kobayashi S, Shimbo T, Fukui T , Cost-effectiveness analysis of screening for osteoporosis inpostmenopausal Japanese women , Journal of Bone and Mineral Metabolism, 2002, 20(6):350-357

Nayak S, Olkin I, Liu H, Grabe M, Gould M K, et al, Meta-analysis: accuracy of quantitative ultra-sound for identifying patients with osteoporosis, Annals of Internal Medicine, 2006, 144(11):832-841

Nelson H D, Morris C D, Kraemer D F, Osteoporosis in postmenopausal women: identification andmonitoring, Agency for Healthcare Research and Quality, 2002, p 320, ISB: 1587630583

Nguyen T.V., Center J.R., Eisman J.A., Bone mineral density-independent association of quantitati-ve ultrasound measurements and fracture risk in women, Osteoporosis International, (2004) 15:12(942-947).

Njeh C.F., Hans D., Li J., et al, Comparison of six calcaneal quantitative ultrasound devices:Precision and hip fracture discrimination, Osteoporosis International, (2000) 11:12 (1051-1062).

Nordstrom A. Gerdhem P. Brandstrom H.et al. Interleukin-6 promoter polymorphism is associatedwith bone quality assessed by calcaneus ultrasound and previous fractures in a cohort of 75-year-old women. Osteoporosis International (2004) 15:10 (820-826).

Omodei U., Benussi C., Ramazzotto F., et al, Risk Identification of Osteoporosis in PostmenopausalWomen by a Simple Algorithm Based on Ultrasound Densitometry and Body Mass Index, Journalof Clinical Densitometry, (2008) 11:3 (412-416).

Pafumi C., Chiarenza M., Zizza G., et al, Role of DEXA and ultrasonometry in the evaluation ofosteoporotic risk in postmenopausal women, Maturitas, (2002) 42:2 (113-117).

Panichkul S, Panichkul P, Sritara C, Tamdee D, Cost-effectiveness analysis of various screeningmethods for osteoporosis in perimenopausal Thai women, Gynecologic and ObstetricInvestigation, 2006, 62(2):89-96

72

Panichkul S. Sripramote M. Sriussawaamorn N. Diagnostic performance of quantitative ultrasoundcalcaneus measurement in case finding for osteoporosis in Thai postmenopausal women. Journalof Obstetrics and Gynaecology Research (2004) 30:6 (418-426).

Pennisi P. Signorelli S.S. Riccobene S. et al Low bone density and abnormal bone turnover in patientswith atherosclerosis of peripheral vessels. Osteoporosis International (2004) 15:5 (389-395).

Pereda C.A., Hannon R.A., Naylor K.E., The impact of subcutaneous oestradiol implants on bioche-mical markers of bone turnover and bone mineral density in postmenopausal women. BJOG (2002)109:7 (812-820).

Pinheiro M.M. Castro C.M. Szejnfeld V.L. et al, Low femoral bone mineral density and quantitativeultrasound are risk factors for new osteoporotic fracture and total and cardiovascular mortality:A 5-year population-based study of Brazilian elderly women. Journals of Gerontology - Series ABiological Sciences and Medical Sciences (2006) 61:2 (196-203).

Pinheiro M.M., Castro C.H.M., Frisoli Jr. A., et al, Discriminatory Ability of Quantitative UltrasoundMeasurements is Similar to Dual-Energy X-Ray Absorptiometry in a Brazilian Women Populationwith Osteoporotic Fracture, Calcified Tissue International, (2003) 73:6 (555-564).

Pluskiewicz W., Tarnawska B., Drozdzowska B., Mandibular bone mineral density measured usingdual-energy X-ray absorptiometry: Relationship to hip bone mineral density and quantitative ultra-sound at calcaneus and hand phalanges, British Journal of Radiology (2000) 73:867 (288-292).

Pongchaiyakul C. Panichkul S. A nomogram for predicting osteoporosis risk based on age, weightand quantitative ultrasound measurement. Osteoporosis International (2007) 18:4 (525-531).

Pongchaiyakul C. Panichkul S. Songpatanasilp T. Combined clinical risk indices with quantitativeultrasound calcaneus measurement for identifying osteoporosis in Thai postmenopausal women.Journal of the Medical Association of Thailand (2007) 90:10 (2016-2023).

Saadi H.F. Reed R.L. Carter A.O. et al. Correlation of quantitative ultrasound parameters of the cal-caneus with bone density of the spine and hip in women with prevalent hypovitaminosis D.Journal of Clinical Densitometry (2004) 7:3 (313-318).

Saleh MM, Jorgensen HL, Lauritzen JB. Odds ratios for hip- and lower forearm fracture using peri-pheral bone densitometry; a case-control study of postmenopausal women. Clin Physiol FunctImaging 2002; 22(1):58-63.

Schneider J, Bundschuh B, Spath C et al. Discrimination of patients with and without vertebral frac-tures as measured by ultrasound and DXA osteodensitometry. Calcif Tissue Int 2004; 74(3):246-54.

Segal E, Dvorkin L, Lavy A et al. Bone density in axial and appendicular skeleton in patients withlactose intolerance: influence of calcium intake and vitamin D status. J Am Coll Nutr 2003;22(3):201-7.

Sennels H.P. Sand J.C. Madsen B. et al. Association between polymorphisms of apolipoprotein E,

73

bone mineral density of the lower forearm, quantitative ultrasound of the calcaneus and osteo-porotic fractures in postmenopausal women with hip or lower forearm fracture. ScandinavianJournal of Clinical and Laboratory Investigation (2003) 63:4 (247-258).

Sim M F, Stone M, Johansen A, Evans W, Cost effectiveness analysis of BMD referral for DXA usingultrasound as a selective pre-screen in a group of women with low trauma Colles' fractures(Structured abstract), Technology and Health Care, 2000, 8(5): 277-284

Sosa M, Saavedra P, Jodar E et al. Bone mineral density and risk of fractures in aging, obesepost-menopausal women with type 2 diabetes. The GIUMO Study. Aging Clin Exp Res 2009;21(1):27-32.

Sosa M., Saavedra P.., Del Pino-Montes J., et al. Postmenopausal women with Colles' fracturehave lower values of bone mineral density than controls as measured by quantitative ultrasoundand densitometry. Journal of Clinical Densitometry (2005) 8:4 (430-435).

Speden D.J., Calin A.I., Ring F.J., et al, Bone mineral density, calcaneal ultrasound, and bone turno-ver markers in women with ankylosing spondylitis, Journal of Rheumatology, (2002) 29:3 (516-521).

Steinschneider M, Hagag P, Rapoport MJ, Weiss M. Discordant effect of body mass index on bonemineral density and speed of sound. BMC Musculoskelet Disord 2003; 4:15.

Stewart A, Kumar V, Reid DM. Long-term fracture prediction by DXA and QUS: a 10-year prospec-tive study. J Bone Miner Res 2006; 21(3):413-8.

Su R.-K. Liu X.- Deng Y.-H.M. Evaluation of the variables changes in quantitative ultrasound mea-surement on osteoporosis-related fracture. Chinese Journal of Clinical Rehabilitation (2005) 9:15(246-247).

Sumino H, Ichikawa S, Kasama S et al. Relationship between brachial arterial endothelial functionand lumbar spine bone mineral density in postmenopausal women. Circ J 2007; 71(10):1555-9.

Sumino H, Ichikawa S, Kasama S et al. Relationship between carotid atherosclerosis and lumbarspine bone mineral density in postmenopausal women. Hypertens Res 2008; 31(6):1191-7.

Takaishi Y. Okamoto Y. Ikeo T. et al. Correlations between periodontitis and loss of mandibularbone in relation to systemic bone changes in postmenopausal Japanese women. OsteoporosisInternational (2005) 16:12 (1875-1882).

Tao B. Liu J.-M. Zhao H.-Y. et al Differences between measurements of bone mineral densities byquantitative ultrasound and dual-energy X-ray absorptiometry in type 2 diabetic postmenopau-sal women. Journal of Clinical Endocrinology and Metabolism (2008) 93:5 (1670-1675).

Tauchmanova A. Nuzzo V.Del Puente L. et ak Reduced bone mass detected by bone quantitativeultrasonometry and DEXA in pre- and postmenopausal women with endogenous subclinical hyper-thyroidism. Maturitas (2004) 48:3 (299-306).

74

Teixeira PJ, Rocha P, Carnero E, Sardinha LB. Relationship between absorptiometry and ultrasoundmeasurements of abdominal subcutaneous fat tissue in postmenopausal women. Acta Diabetol2003; 40 Suppl 1:S76-8.

Toussirot E., Michel F., Wendling D., Bone density, ultrasound measurements and body composi-tion in early ankylosing spondylitis, Rheumatology (2001) 40:8 (882-888).

Tremollieres FA, Bauvin E, Cigagna F et al. Association of cardiovascular risk factors with intima-media thickness of the carotid arteries in early postmenopausal women. Menopause 2004;11(3):323-30.

Tsuritani I, Brooke-Wavell KS, Mastana SS, Jones PR, Hardman AE, Yamada Y. Does vitamin Dreceptor polymorphism influence the response of bone to brisk walking in postmenopausalwomen? Horm Res 1998; 50(6):315-9.

Tuzun S. Tangurek S. Devranoglu G. et al. Preoperative Evaluation of Bone Mass Using Dual X-RayAbsorptiometry and Quantitative Ultrasound in Postmenopausal Women with PrimaryHyperparathyroidism European Surgery - Acta Chirurgica Austriaca (2003) 35:4 (207-210).

Winzenberg Tania M, Shaw Kelly A, Fryer Jayne, Jones Graeme, Calcium supplementation for impro-ving bone mineral density in children, Calcium supplementation for improving bone mineral densi-ty in children, Cochrane Database of Systematic Reviews: Reviews 2006 Issue 2, John Wiley & Sons,Ltd Chichester, UK DOI: 10.1002/14651858.CD005119.pub2, 2006, 2, John Wiley & Sons, Ltd

75

77

Appendix 5

Tables of evidenceStudy Id(design) Objective Population Intervention Comparator Outcome Quality

(QUADAS items) Notes

Clowes2006(cohort,UK)

To define 95%chance of detect-ing osteoporosiswith peripheral(only QUS dataextracted) devices(expressed as aROC curve with a95% cuttoff) com-pared to DEXA –hip BMD

3 non-randomlyrecruited popula-tions:

-500 55-80 yearspostm in wedgesof 100 for each 5-year span (includ-ing 279 fracturecases and genpopulation)

-100 prem 20-40years (“youngnormal popula-tion”)

DEXA lumbarspine

(L1–L4) & proxfemur: AcclaimQDR

4500 (HologicUSA)

QUS calcaneus:Achilles + (GELunar, USA) andUBIS 5000

(DiagnosticMedicalSystems,France) andDTUOne(OsteometerMeditech, USA)and QUS 2 (QuidelMetra, USA). QUSradius, metatarsal,3rd phalanx:Omnisense(Sunlight, Israel)QUS proximalphalanges: DBMSonic BP (Igea,Italy)

Peripheral QCTdistal forearm XCT2000 (Stratec,Pforzheim,Germany):

DTX 200 distalforearm (OsteometerMeditech,Hawthorn, Calif,USA

T scores and cor-relation arereported for thepopulation cohortnot broken downby fracture status.However, Table 3and 4 report bydevice and site ofmeasurement ahigh sensibilityand low specificity

Out of 12 applica-ble items, 8 Yes 2unclear and 2 par-tial items. Overallgood study

There is widelydifferent perform-ance betweendevices and acrosspopulations andthe authors pro-pose using twothresholds with95% sensitivity orspecificity to iden-tify people whoneed or do notneed treatment orfurther investiga-tion

Dane 2008(cohort,Turkey)

To assess correla-tion betweenDEXA and QUSand calculate QUSop predictivepower

175 prem and 202postm aged 29-73. 26 doputs notincluded in theanalysis

DEXA femoralneck and spineL2-L4: HologicQDR 4500

QUS calcaneusLunar AchillesExpress parame-ters: BUA, SOSand stiffness index(SI)

Expressed as BUA,SOS and SI forQUS and trans-formed into Tscores and plottedas BMD into ROCcurves. All resultsreported separate-ly for prem andpostm includingPPV and NPV

Out of 14 relevantitems 3 Yes, 2 No,1 Partly and 8unclear. The studyis well reported

The authors con-clude that sensi-tivity and specifici-ty where not highenough for QUS tobe used in alter-native to DEXA.The study results,although poorlyreported, are con-sistent with theother evidence ofreasonable sensi-tivity and poorspecificity of QUScompared to DEXA

78

Study Id(design) Objective Population Intervention Comparator Outcome Quality


Frost 2001(cohort, UK)

To compare per-formance of DEXAvs QUS and cor-relate risk factorsto tests’ perform-ance

1115 prem andpostm with andwithout risk fac-tors. Data extract-ed only for 146untreated postmwithout risk fac-tors aged 59.9(8.4), BMI of 25(3.5)

DEXA lumbarspine (L1 – L4)and hip(neck,total) BMD.QDR4500(Hologic, USA)

QUS (calcaneuswith estimation ofheel BMD) param-eters: BUA ansSOS QUS HologicSahara andDTUone(OsteometerMediTech, USA)

Z and T scores(measurementvalue standardizedby populationage). Multipleregression for Z, Tand risk factors.

Out of 13 applica-ble items, 9 Yes 2unclear and 2 par-tial items. Overallgood study

The authors con-clude that QUS issensitive but pro-vides 0.5 T scoreslower than DEXA.Z scores areequivalent. Thisimplies a lowerspecificity if usedwith the currentWHO criteria.With the revisedWHO criteria itsperformance isthe same as DEXAbut more portableand with less radi-ation. Theauthors’ sugges-tion that WHOclassificationthresholds bechanged needs tobe interpretedwith caution

Jemalmaz2007(cohort,Turkey)

Comparison of per-formance of QUSvs DEXA

116 postm report-ing consecutivelyto OPD aged 57.3(8.4) years. 19%with a history offractures (data notreported separate-ly).

DEXA femoral neckand L1-L4: QDR-4500 (Hologic,USA)

QUS left heel:Achilles Express(Lunar, USA). Nodata about QUSparameters

BMD and T scorefor DEXA. SI-derived T score forQUS (op as perWHO guidelines)

Out of 12 relevantitems 3 Yes, 2 No,1 unclear. Thestudy is vaguelyreported

The authors con-clude that QUS canbe used for strati-fying

the population intorisk groups (DEXAfor those T-score -2.2). This conclu-sion should besubject to econom-ic evaluation

Larijani B. etal (Tehran,Iran, 2005)Cross sec-tional study

To seek the propercutoff value forQUS of the heel inidentifying BMDcategories asdetermined byDEXA in post-menopausalwomen.

420 healthy post-menopausalwomen. Age range45-80 years;155+-7 cm. forHeight; 66+-11Kg. for weight;26+-6 for BMI;Mean menopausalduration: 10+-8.

DEXA with lunarinstrument (Lunarcorporation) atlumbar spine (L2-L4) and left femur(neck, trochanter,total).

QUS of the left cal-caneus (Achilles:Lunar corporation)expressed as BUA,SOS and SI.No preference forleft or right

Prevalence ofosteoporosis andosteopenia accord-ing to DXA;Specificity andsensitivity andPPV% and NPV%of QUS comparedwith DXA at eachregion. T scoreand 99.5% CI forQUS inOsteoporoticParticipants(according toDXA). AUC

Out of relevantitems 1 No, 4unclear. Study. Fairreported study.

Authors concludethat there is insuf-ficient agreementbetween QUS andDXA and this ledto uncertainty onexpected BMD inpeople tested withQUS. The pro-posed cutoff valuecould achievehigher sensitivitybut only by accept-ing higher rates offalse positiveresults.

Key: BMD = bone mineral density; SOS = speed of sound; BUA = broadband US attenuation; AUC = areaunder the curve; op = osteoporosis; prem = premenopausal women; postm = postmenopausal women; SI =stiffness index; pt = patient;

79

Study Id(design) Objective Population Intervention Comparator Outcome Quality


MacLaughlin2005(cohort, USA)

To identify pt atrisk of op andcompare per-formance of QUSvs DEXA in com-munity setting

97 mainly postmaged mean 66.2(7.9) BMI mean28.47 (6.24)

DEXA (No dataabout useddevice) on ptwith T score < -1at L1-4, bothfemur (neck) andthrocanters andleft and rightforearm

QUS non-domi-nant heel:Achilles Express(Lunar, USA).Stiffness index(SI)

SI-derived Tscore for QUS

Out of 14 rele-vant items 4 Noand 1 partly.Badly designedand reportedstudy

This is a commu-nity screeningstudy. Its conclu-sions that QUS isuseful to detectrisk cases shouldbe regarded withcaution in thelight of studyquality and smallnumbers

Scott 2004(cohort, EPIDOSstudy, France)

To compare thepredictive powerof BMD with QUSmeasurement invery elderlywomen

7598 womenaged > 75. 5978women only hadQUS assessment(230 fracturedwomen during3.5 years of fol-low up). 538 lostto follow-up

DEXA BMD prox-imal femur: DPXPlus (GE Lunar,USA)

QUS calcaneus:Achilles system(GE-Lunar, USA).Expressed inSOS, and BUA

Expressed asage-stratifiedBMD ROC curves(AUC) and risk offracture by multi-ple regression

Out of 14 rele-vant items 2 No,2 unclear and 1partly. Complex,well designedand well reportedstudy

The authors con-clude that sensi-tivity and speci-ficity of BMDdecrease withage, but BMD is abetter predictorof fracture up toage 80 than SOSand BUA. Thestudy does notfocus on theidentification ofop per se, only ofprediction of hipfractures

81

Appendix 6QUADAS for included studies

Clowes JA, Peel NF, Eastell R.Device-specific thresholds to diagnose osteoporosis at the proxi-mal femur: an approach to interpreting peripheral bone measurements in clinicalpractice.Osteoporos Int. 2006;17(9):1293-302.

82

Dane, C., Dane, B., Cetin, A. and Erginbas, M.(2008) 'The role of quantitative ultrasound inpredicting osteoporosis defined by dual-energy X-ray absorptiometry in pre- and postmenopausalwomen',Climacteric,11:4,296 - 303.

Frost, M. L., Blake, G. M., Fogelman, I. Quantitative Ultrasound and Bone Mineral Density AreEqually Strongly Associated with Risk Factors for Osteoporosis Journal of bone and mineral rese-arch Volume 16, Number 2, 2001 p.406-416.

83

Gemalmaz A., Güzel D., Ceylan C. Diagnostic Performance of QUS for Identifying Osteoporosisin Postmenopausal Turkish Women. Turk J Med Sci 2007; 37 (5): 303-309

84

Larijani B., M. Dabbaghmanesh H., Aghakhani S., et al. Correlation of Quantitative HeelUltrasonography with central dual-energy X-ray Absorptiometric bone mineral density in postme-nopausal women. Ultrasound Med 2005 24:941-946

85

MacLaughlin E.J., MacLaughlin A.A., Snella K.A., et al. Osteoporosis Screening and Educationin Community Pharmacies using a team approach. Pharmacotherapy 2005; 25(3):379-386

86

Schott A.M., Kassaý¨ Koupaý¨ Æ B., Hans D., et al. Should age influence the choice of quan-titative bone assessment technique in elderly women? The EPIDOS study. Osteoporos Int (2004)15: 196-203

87

89

Appendix 7

Survey questionnaire

The following questionnaire was for centres with both DEXA and QUS. Centres performing onlyDEXA or only QUS were sent a shorter instrument including only the part of the below question-naire the are about DEXA or QUS.

101

Appendix 8

Responding centres

Key: P.A. = autonomous province (regional equivalent)

Region Responding centre

P.A. Bolzano Ospedale di Bolzano

Piemonte Amb. Ospedale di Aquiterme

Lombardia Ospedale L. Sacco

Piemonte A.O.U. Molinette

Piemonte Osp. Ostetrico Gin. Sant’Anna

Valle d’Aosta Presidio osp. di Viale Ginevra

Friuli Venezia Giulia A.O. Santa Maria degli Angeli

Marche Osp. Santa Maria della pietà

Veneto A.O. di Verona Poliambulatorio

Emilia Romagna A.O.U. Policlinico S. Orsola

Puglia Osp. Casa sollievo della sofferenza

Veneto Poliamo. Po De Gironcoli

Emilia Romagna Amb. AUSL di Imola

Lazio Osp. Riuniti Anzio e Nettuno

Toscana Osp. di Santa Chiara (A.O. Pisana)

103

Acronyms and abbreviations

Agenas Agenzia nazionale per i servizi sanitari regionali

BMD Bone Mineral Density

BUA Broadband ultrasound attenuation expressed as dB/MHz

CE Mark Conformité Européenne Mark

CL Cochrane Library

CUD Commissione Unica Dispositivi medici (Ministerial Committee on MedicalDevices)

DEXA Dual-energy x-ray absorptiometry

DRG Diagnosis related group

ICD-9-CM International Classification of Diseases, Ninth Revision, ClinicalModification

ICER Incremental cost-effectiveness ratio

INHS Italian National Health Service

ISTAT National Institute of Statistics

MLHSP Ministry of Labour, Health and Social Policy

MRI Magnetic Resonance Imaging

P.A. Provincie Autonome, regional equivalent

PQ Policy Question

PT Patients

QCT Quantitative Computed Tomography

QUADAS Quality Assessment of Diagnostic Accuracy Studies

QUS Quantitative Ultrasound

RA Radiographic Absorptiometry

RCT Randomised Controlled Trial

SDO Scheda di dimissione ospedaliera (hospital discharge record)

SI Stiffness index

SOS Speed of sound expressed as m/s

SPA and DPA Single and Double Photon Absorptiometry

VAT Value Added Tax

VC Variation coefficient

WHO World Health Organisation

age.na.s - Agenzia Nazionale per i Servizi Sanitari Regionali

Sezione A.G.P. - Centro Stampa

Via Puglie 23, 00187 – Roma . Tel. 06.427491 – fax. 06.42749488

www.agenas.it e-mail [email protected]

hta dexa intro€¦ · This report should be cited as: Amicosante AMV, Bernardini F, Cavallo A,...

Documents

Transcript of hta dexa intro€¦ · This report should be cited as: Amicosante AMV, Bernardini F, Cavallo A,...