Exploring Patients' Perception of Osteoporosis Following a ...

Exploring Patients' Perception of Osteoporosis Following a Fragility Fracture: Results of a Literature Review and

Analysis of a Provincial Database

by

Rebeka Sujic

A thesis submitted in conformity with the requirements for the degree of Master of Science

Department of Health Policy, Management and Evaluation University of Toronto

© Copyright by Rebeka Sujic, 2011

ii

Exploring Patients' Perception of Osteoporosis Following a

Fragility Fracture: Results of a Literature Review and Analysis of

a Provincial Database

Rebeka Sujic

Master of Science

Department of Health Policy, Management and Evaluation University of Toronto

2011

Abstract

Many patient-level barriers to more effective post-fracture osteoporosis (OP) management are

associated with patients’ perceptions of their bone health and the link between their fracture and

OP. These barriers could be addressed by interventions which account for the role of patients’

perceptions, such as those based on theories of behaviour change. This thesis had two objectives:

to review the literature and determine whether these theories have been integrated in post-

fracture OP management and to examine patients’ awareness of the OP-fracture link. The results

showed that theories of behaviour change have not been integrated into post-fracture

interventions, that most patients do not perceive the OP-fracture link even after a baseline

intervention and that two baseline factors predict who is less likely to make the OP-fracture link

at follow up. Based on these findings, modifications to current OP interventions were proposed

and guidance that could help create new, patient-centered interventions was provided.

iii

Acknowledgments I would first like to acknowledge my thesis committee: my supervisor, Dorcas Beaton, PhD,

without whom this thesis would not have been possible, and my thesis committee members,

Monique Gignac, PhD and Rhonda Cockerill, PhD for their invaluable time, feedback and

support.

I am also very grateful to Osteoporosis Canada, particularly Mr Ravi Jain, for providing me with

access to a one-of-a-kind source of data and the Osteoporosis Screening Coordinators and Area

Managers who have all been a pleasure to work with. I would like to thank the Ministry of

Health and Long Term Care of Ontario and St Michael’s hospital Foundation (Sister Christine

Gaudet Scholarship Fund and Dr. S. Gordon Ross Memorial Fund) for the financial support.

Everyone at the Mobility Program Clinical Research Unit who provided me with the

encouragement and help – thank you. I would also like to thank my examiners, Dr. Deborah

DaCosta and Dr. Curtis Breslin.

There are several people and one animal that I owe special thanks to: my friend Chan who

helped edit my thesis drafts often on very short notice; my supportive aunt who was a great

source of motivation for me to finish the thesis; my late grandmother who was always proud of

me and who had announced to everyone that I had a PhD before I started my master’s; my

friends, who did not disown me and did not forget me even though they saw me rarely during the

past two years; and my cat Az, who took his role as a paper weight very seriously and would

regularly sit on papers scattered around me, particularly the ones that I needed.

Everyone else who I forgot to mention here and who I will likely remember once this thesis is in

print – thank you!

Disclaimer: This project was funded by the Ministry of Health and Long Term Care of Ontario.

The views expressed are those of the researcher and do not necessarily reflect those of the

Ministry.

iv

iv

vi

vii

viii

1

1

1

2

5

6

7

8

10

10

13

15

17

22

22

Table of Contents

Abstract ...........................................................................................................................................ii

Acknowledgments.......................................................................................................................... iii

Table of Contents...........................................................................................................................

List of Tables .................................................................................................................................

List of Figures ...............................................................................................................................

List of Appendices .......................................................................................................................

Chapter 1 Introduction and Objectives ...........................................................................................

1.1 Osteoporosis and fragility fractures ....................................................................................

1.2 Burden of osteoporosis and fragility fractures....................................................................

1.3 The importance of osteoporosis management after a fragility fracture ..............................

1.4 Current post-fracture osteoporosis management: gap in care............................................. 3

1.5 Barriers to more effective osteoporosis management .........................................................

1.6 Patient-level barriers to improving osteoporosis management ...........................................

1.7 Objectives ...........................................................................................................................

1.8 Chapter summary ................................................................................................................

Chapter 2 A Review of Patient-centered Post-fracture Osteoporosis Interventions in the Context of Health Behaviour Change Theories .......................................................................

2.1 Introduction.......................................................................................................................

2.2 Methods.............................................................................................................................

2.3 Results...............................................................................................................................

2.4 Discussion .........................................................................................................................

Chapter 3 Critical Appraisal of the Ontario Osteoporosis Strategy’s Fracture Clinic Screening Program Database as a Source of Data for the Thesis ............................................

3.1 Introduction.......................................................................................................................

v

22

25

26

34

36

38

40

40

42

46

50

54

54

56

57

60

61

63

65

3.2 Background and description of the database ....................................................................

3.3 Critique of the database: establishing its suitability to the research question...................

3.3.1 Purpose and content of the database, sampling, comparison of the sample to the target population and data collection methods............................................

3.3.2 Database structure, content of each observation, duplicates and coding.................

3.3. 3 Manual inspection of data for data flaws................................................................

3.4. Chapter summary ..............................................................................................................

Chapter 4 Factors Associated with Patients Making the Link Between Their Fragility Fracture and Osteoporosis........................................................................................................

4.1 Introduction.......................................................................................................................

4.2 Methods..............................................................................................................................

4.3 Results ................................................................................................................................

4.4 Discussion ..........................................................................................................................

Chapter 5 Discussion and Conclusion ..........................................................................................

5.1 Discussion ..........................................................................................................................

5.2 Key findings .......................................................................................................................

5.3 Implications........................................................................................................................

5.4 Strengths.............................................................................................................................

5.5 Limitations .........................................................................................................................

5.6 Areas for future research....................................................................................................

References.....................................................................................................................................

Appendices.................................................................................................................................... 80

vi

List of Tables

Table 2.1 Main concepts of five major individual-level theories of health

behaviour change ……………………………………………………..…12

Table 3.1 Description and comparison of the sample (n=853) to the target

population (n=2237) ……………………………………………….……32

Table 3.2 Manual examination of raw data: missing values in three data

samples…………… ……………………………………………….……37

Table 4.1 Regression of baseline factors most likely to be associated with a

transition to making the osteoporosis-fracture link…………….…..……49

vii

List of Figures

Figure 2.1 Search strategy for the literature review of post-fracture osteoporosis

interventions …………………………………………………….………14

Figure 3.1 Patient flow through the Fracture Clinic Screening Program……………24

Figure 3.2 Sampling strategy for the osteoporosis-fracture link analysis …………..29

Figure 3.3 Comparison of the sample (n=853) to the target population (n=2237)….31

Figure 4.1 Hypothesized predictors of making the osteoporosis-fracture link by

follow up ………………………………………………………………..44

Figure 4.2 Follow up distribution of those who did not make the osteoporosis-

fracture link at baseline………………………………………………......47

viii

List of Appendices

Appendix I Description of the studies included in the literature review of

patient-centered post-fracture osteoporosis interventions in the

context of theories of behaviour change…………………………..….….80

Appendix II Fracture Clinic Screening Program’s patient educational materials

and Program Information Letter .………………………………………..86

Appendix III Construct list with description of all variables collected as part of the

Fracture Clinic Screening Program ……………………………………...88

Appendix IV Research ethics approval ………………………………………….……..93

1

Chapter 1 Introduction and Objectives

1.1 Osteoporosis and fragility fractures Osteoporosis (OP) is a chronic, asymptomatic disease characterized by a decrease in bone

density, making bones fragile and susceptible to fragility fractures (NIH Consensus Development

Conference on Osteoporosis Prevention, 2000). The prevalence of OP increases with age and,

depending on the fracture site where bone density is measured, ranges from 6-14.8% for women

aged 50-59 to 41-70% for those over the age of 80 (Kmetic, Joseph, Berger, & Tenenhouse,

2002; Melton, III, 1995). Fragility fractures occur as a result of low trauma, typically falling

from standing height or less (Brown & Josse, 2002) and are defined by the World Health

Organization (1998) as fractures “…caused by injury that would be insufficient to fracture

normal bone: the result of reduced compressive and /or torsional strength of bone”. Patients who

sustain one fragility fracture are at 1.5 to 9.5 fold higher risk of future fractures (Klotzbuecher,

Ross, Landsman, Abbott, III, & Berger, 2000).

OP and related fractures are a major global health problem: it is estimated that 9 million

osteoporotic fractures occurred worldwide in the year 2000 alone (Kanis & Johnell, 2005). The

estimates from the US show that 10 million Americans over the age of 50 have OP and that

another 34 million are at risk of the disease (NIH Consensus Development Conference on

Osteoporosis Prevention, 2000; U.S. Department of Health and Human Services, 2004). In

Canada, OP affects 16% of women and 6.6% of men over the age of 50 (Tenenhouse et al.,

2000). According to the estimates from the UK, one of two women and one of five men over the

age of 50 will have an osteoporotic fracture in their lifetime (van Staa, Dennison, Leufkens, &

Cooper, 2001).

1.2 Burden of osteoporosis and fragility fractures OP and resulting fractures are associated with high human and health care costs. Fractures are

associated with a decrease in patients’ mobility, independence, quality of life (Papaioannou et

2

al., 2002; Adachi et al., 2002) and, in case of vertebral fractures, severe kyphosis with pain, loss

of height, depression and low self-esteem (Hawker, 1996; Gold, 1996). Hip fractures are a major

cause of disability and death: one year after sustaining a hip fracture 50% of patients become

functionally dependent in their everyday activities, 40% of patients become unable to walk

independently, and up to 27% enter a nursing home as a result of the fracture (Papaioannou,

Clarke, Campbell, & Bedard, 2000; Wiktorowicz, Goeree, Papaioannou, Adachi, &

Papadimitropoulos, 2001). Mortality associated with hip fractures is particularly high:

approximately 20% of women and 40% of men die within one year of fracturing their hip

(Papaioannou et al., 2000; Wiktorowicz et al., 2001; Bierman AS, 2010).

Health care costs of OP and related fractures are high both in Canada and internationally. In

Canada, the estimated health care costs of treating OP and fractures are expected to rise from

estimated $1.3 billion in 1993 to $10-$20 billion annually (Lorrain et al., 2003; Goeree &

O'Brien B, 1996). Despite literature suggesting a decline in the incidence of hip fractures (Jaglal

et al., 2005), the burden will still be considerable. In Canada, the annual economic burden of hip

fractures alone is projected to increase to $2.4 billion by 2041 (Wiktorowicz et al., 2001).

European estimates from 2000 show that healthcare costs for 3.8 million osteoporotic fractures

totalled 31.7 billion Euros (Kanis & Johnell, 2005). Based on fracture probability and changing

demographic factors in Europe, this figure is expected to increase to 76.7 billion Euros in 2050

(Kanis & Johnell, 2005).

1.3 The importance of osteoporosis management after a fragility fracture

The importance of assessing fragility fracture patients for OP has been emphasized in both

previous research (Astrand, Thorngren, & Tagil, 2006; Lindsay et al., 2001; Center, Bliuc,

Nguyen, & Eisman, 2007; Astrand et al., 2006) and guidelines for OP management (Brown et

al., 2006; Brown & Josse, 2002; Kanis et al., 2008; DH/SC, 2009; National Osteoporosis

Foundation, 2010; van Staa et al., 2001). Previous research found that screening fragility fracture

patients is an effective way of identifying those with a low bone mineral density (BMD) who

should be assessed for OP: one study found that 87% of fragility fracture patients in orthopaedic

environment had BMD that was at least one SD below the mean (Astrand et al., 2006). A large

3

OP epidemiological study of community-dwelling men and women in Australia found that about

50% of fragility fracture patients experienced another fracture in the 10 years following the

initial fracture (Center et al., 2007). Because of the large number of patients who re-fractured,

the study concluded that even one fragility fracture is a clear indication for therapy aimed at

prevention of future fractures. An international study analysing data from 373 study centers in

North America, Europe, Australia, and New Zealand found that 20% of post-menopausal women

with vertebral fractures sustained another fracture within a year of the first fracture (Lindsay et

al., 2001). These findings suggest that urgent action needs to be taken to prevent future fractures

particularly because they may occur a lot sooner than expected.

In addition to previous research, both Canadian and international guidelines emphasize the need

for OP management and prevention of future fractures in fragility fracture patients (Brown &

Josse, 2002; Brown et al., 2006; 1998; DH/SC, 2009; Guideline Development Group for the

Royal College of Physicians, 2010; National Osteoporosis Foundation, 2010; NIH Consensus

Development Conference on Osteoporosis Prevention, 2000; U.S.Department of Health and

Human Services, 2004; National Institute for Health and Clinical Excellence., 2008). According

to the Canadian guidelines, a previous fragility fracture is one of the four most significant risk

factors for future fractures and a major risk factor for identifying who should be assessed for OP

(Brown & Josse, 2002; Brown et al., 2006). 2008 US guidelines recommend testing in all adults

with fragility fractures and considering pharmacotherapy for those over the age of 50 who have

sustained a hip or vertebral fracture and those with a low BMD who have a high probability of

future fractures (National Institute for Health and Clinical Excellence., 2008). 2002 UK Royal

College of Physicians’ guidelines recommend the that bone protective therapy be considered for

all fragility fracture patients over the age of 64 while the National Institute for Health and

Clinical Excellence recommends that female fragility fracture patients over the age of 75 should

receive treatment even without undergoing a diagnostic test (National Institute for Health and

Clinical Excellence, 2008).

1.4 Current post-fracture osteoporosis management: gap in care Effective treatment options for OP are available. Medication can reduce fracture risk by 30- 50%

among those who have OP (or a previous vertebral fracture) (MacLean et al., 2008; Cranney et

4

al., 2002). Bisphosphonates, a class of drugs which inhibits bone resorption and therefore

prevents the loss of bone mass, have been shown to decrease the risk of fractures: the relative

risk of fracture for women using bisphosphonates (vs. placebo) is 0.47 for hip, 0.52 for vertebral

and 0.70 for other fractures (Black & Cooper, 2000). Both Canadian and international guidelines

support the use of such pharmacological agents in reducing the risk of future fractures (Kanis et

al., 2008; Brown & Josse, 2002; Brown et al., 2006; Lewiecki & Watts, 2009). Canadian Clinical

Practice Guidelines, for example, advise that bisphosphonates be used as treatment to decrease

vertebral, non-vertebral and hip fractures and recommend Calcium and vitamin D as essential for

prevention of OP and as an adjunct to the main pharmacotherapy for OP treatment (Brown &

Josse, 2002; Brown et al., 2006).

Despite the existence of effective treatment options and clinical practice guidelines

recommending that fragility fracture patients be assessed (and treated if appropriate) for OP, the

majority of fragility fracture patients do not receive appropriate OP management. A systematic

review examining OP management of fragility fracture patients in the US, Canada, Europe,

Australia and New Zealand found that these patients were not commonly tested and treated for

OP: 1- 32% of patients had diagnostic testing (BMD test) done while pharmacotherapy was

reported in 1% to 65% of patients (Giangregorio, Papaioannou, Cranney, Zytaruk, & Adachi,

2006). Another systematic review of diagnostic and treatment patterns of OP after a fragility

fracture similarly found that less than 32% of patients underwent BMD testing as part of OP

management, even though 42% to 100% of tested patients ended up diagnosed with OP (Elliot-

Gibson, Bogoch, Jamal, & Beaton, 2004). This review also showed that only 0.5–38% of patients

who had a BMD test were prescribed bisphosphonates, while Calcium and Vitamin D were used

in only 8–62% of cases (median 18%). A large UK audit in 172 trusts confirmed the low rate of

post-fracture testing and treatment: despite a number of government-led initiatives that aim at

improving the care of fragility fracture patients, post-fracture assessment and treatment rates

remained low with only 20-35% of older female fragility fracture patients assessed for OP and

only 19% of those with non-hip fractures prescribed pharmacotherapy (Clinical Effectiveness

and Evaluation Unit.London, 2007).

5

1.5 Barriers to more effective osteoporosis management Barriers to more effective OP management are associated with both the healthcare environment

and the patient. Barriers associated with the healthcare environment include a lack of OP

knowledge among physicians (such as belief that OP is a normal part of aging with a low

mortality rate), existence of competing conditions that take priority over discussing OP during a

medical visit / an insufficient amount of time that physicians may have to address OP, a lack of a

consensus on which health professional is responsible for OP management, a lack of access to

diagnostic testing as well as the complexity in coordinating patient identification, testing and

treatment initiation (Simonelli, Killeen, Mehle, & Swanson, 2002; Kaufman et al., 2003; Bogoch

et al., 2006).

A number of interventions have been implemented to address these barriers including physician-

targeted education and reminders (Bliuc, Eisman, & Center, 2006; Majumdar et al., 2004; Jaglal

et al., 2009), appointing a healthcare provider or a multidisciplinary healthcare team to address

OP issues (Chevalley, Hoffmeyer, Bonjour, & Rizzoli, 2002; Ho, Cranney, & Campbell, 2006;

Levasseur et al., 2007), as well as creation of complex programs where a dedicated coordinator

facilitates diagnostic testing and coordinates post-fracture OP care (Bogoch et al., 2006;

Majumdar et al., 2007; Levasseur et al., 2007).

Even though addressing and removing healthcare-level barriers is an important step in improving

OP management, it is ultimately the patient who needs to implement the recommendations and

adhere to them. Without the commitment of the patient, interventions are not likely to be

successful: even in cases where specific interventions to improve OP management were

implemented, up to 79% of patients with OP or osteopenia did not initiate treatment (Bliuc et al.,

2006; Chevalley et al., 2002; Majumdar et al., 2004). Of those who initiated treatment, 74%

ceased to take their medication within one year (Edwards et al., 2005). A study involving

patients with low trauma distal forearm fractures revealed that patients may remain resistant to

OP diagnosis and treatment despite the increased OP interventions involving both orthopaedists

and primary care physicians (Cuddihy et al., 2004). The authors of this study suggested that

additional efforts from health care professionals are just one part of the puzzle and that patient-

level barriers to OP treatment need to be examined more closely.

6

1.6 Patient-level barriers to improving osteoporosis management Patient-level barriers to more effective OP management that have been identified include a lack

of perceived susceptibility to OP and future fractures (Meadows, Mrkonjic, & Lagendyk, 2005;

Giangregorio et al., 2008; Burgener et al., 2005), not making the link between OP and fractures

(Meadows 2005, Giangregorio 2008), a lack of readiness to accept OP therapy (Mauck et al.,

2002) and perception that treatment is ineffective or that no treatment is necessary following a

healed fracture (Kaufman et al., 2003). Considering these barriers, it is not surprising that many

patients are reluctant to undergo further testing and initiate treatment: if a patient believes that

OP management is not applicable to them or that the existing treatments are ineffective, they are

not likely to initiate and adhere to the prescribed therapy and recommendations. Meadows

(2003), for example, found that female fragility fracture patients who perceived their fractures as

accidents, rather than markers of poor bone health, took little or no steps towards OP

management.

Luckily, these patient-level barriers are modifiable. They could be modified through

interventions that focus on changing patients’ perception of OP-related issues with the goal of

ultimately changing patients’ health behaviour in favour of initiation of OP management. One

does not have to search far to find information on health behaviour change. The process by

which individuals change their behaviours has been the focus of a large area of research:

individual theories of health behaviour change. Different theories propose different models of

behaviour change but they all concentrate on an individual’s perception of his or her

health/condition (Glanz, Rimer, & Lewis, 2002). Given that many of the barriers to more

effective OP management stem from patients’ perception of their condition, it would be fitting to

apply these theories to the context of post-fracture OP management. Whether these theories have

been applied to the context of post-fracture OP management (and what the resulting outcomes

are) has not been examined so far.

An important barrier to more effective OP management and prevention of future fractures

involves patients’ lack of recognition that their fragility fracture may be associated with an

underlying bone health disease and therefore with the risk of future fractures. Even though a

fragility fracture is the best predictor of future fractures (Klotzbuecher et al., 2000) and the main

risk factor that identifies who should be assessed for OP (Brown et al., 2006), many fragility

7

fracture patients do not make the link between their fracture and OP (Giangregorio et al., 2008).

Making this link is important because a fragility fracture is usually the first and, with the

exception of vertebral fractures, the only “symptom” of OP. The importance of making this link

has been recognized and incorporated into interventions by two domestic, province-wide post-

fracture OP screening initiatives: Recognizing Osteoporosis and its Consequences in Quebec

(ROCQ) (Bessette et al., 2008), a health-management program aimed at evaluation of OP care

following a fragility fracture in Quebec, and the Ontario Osteoporosis Strategy’s Fracture Clinic

Screening Program, a province-wide initiative of the Ministry of Health and Long Term Care of

Ontario (Jaglal et al., 2010). Promoting the awareness of the OP-fracture link is a part of the

intervention implemented by ROCQ, while increasing patient’s awareness of the OP-fracture

link is one of the goals of the Ontario’s Fracture Clinic Screening Program. The two initiatives

emphasize the OP-fracture link for a good reason: if patients perceive that their fracture is due to

an isolated accident, rather than poor bone health, they are not likely to take action following the

fracture as they are not likely to perceive OP management as relevant to them. Perception of

one’s fracture as osteoporotic has been identified as a predictor of follow up with a physician for

investigation of OP (Bliuc et al., 2006) and initiation of OP management (Beaton, Jiang, Sujic,

Sale, & Bogoch, 2009; Beaton, Bogoch, Sujic, & Elliot-Gibson, 2007; Meadows et al., 2005).

Knowing which patients are less likely to make the link (even after a baseline intervention)

would help target interventions to that group of patients in order to help them link OP and their

fracture. The predictors of which undiagnosed and untreated patients are more or less likely to

make the OP-link after a fragility fracture have not yet been examined.

1.7 Objectives The overall purpose of this thesis is to examine the role of patients’ perception of OP in the

context of post-fracture care in order to contribute to the body of knowledge necessary for

designing more effective OP interventions. There are two specific objectives:

1. To determine whether current OP interventions (that aim to improve post-fracture care)

account for patient perceptions of their health and OP. Patients’ perceptions are an important

factor in behaviour change and are a main building block of individual-level theories of

health behaviour change that aim to explain how and why behaviour change occurs. It is

8

unknown whether these theories have been used in the context of post-fracture OP

interventions. My first objective will therefore consist of performing a literature review to

examine whether individual-level theories of health behaviour change have been applied to

post-fracture OP interventions and if so, what were the outcomes. This objective will be

addressed in Chapter 2 (manuscript “A Review of Patient-centered Post-fracture

Osteoporosis Interventions in the Context of Health Behaviour Change Theories”).

2. To identify factors predictive of patients’ making the link between their fracture and OP.

Making the OP-fracture link is important in initiation of post-fracture OP management as

fractures are usually the first signs of an underlying bone health issue and are strongly

associated with future fractures. Factors that predict which patients (who were not previously

diagnosed or treated for OP) are more or less likely to make this link are unknown.

Identification of patients less likely to make the link (after a baseline intervention) and their

characteristics would help create interventions targeted to that group of patients. The second

objective of this thesis is to identify and examine predictors of making the OP-fracture link in

previously undiagnosed and untreated patients. Chapter 4 (manuscript “Factors Associated

with Patients Making the Link Between Their Fragility Fracture and Osteoporosis”) will

focus on this objective.

1.8 Chapter summary OP and related fractures are a global health problem associated with high human and health care

costs. Previous fragility fracture is a major risk factor for future fractures and a major indication

for OP assessment. Despite effective treatment options, guidelines and previous research

advocating assessment and treatment of fragility fracture patients, the majority of patients do not

receive appropriate OP management. Barriers to more effective OP management are associated

with both the healthcare environment and the patient. Removing healthcare-level barriers is

important, however, interventions are not likely to be successful if patient-level barriers are not

addressed. Patient level-barriers are modifiable through interventions focusing on changing

patients’ perception of their condition with a goal of changing their behaviour in favour of

initiating OP management. Theories of health behaviour change would be applicable to this

9

context as they focus on patient perception and aim to explain how and why behaviour change

occurs. The first objective of the thesis is to examine whether theories of health behaviour

change have been applied to post-fracture interventions. The second objective will focus on

patient perception of the OP-fracture link as many patients do not make this link. Making the

OP-fracture link is important because a fragility fracture is usually the first and only sign of

deteriorating bone health. The second objective will examine the predictors of making and not

making this link in undiagnosed and untreated patients in order to identify the patients who are

not likely to make the link and who may require a more intensive intervention.

10

Chapter 2 A Review of Patient-centered Post-fracture Osteoporosis

Interventions in the Context of Health Behaviour Change Theories

2.1 Introduction OP is a chronic disease characterized by a decrease in bone density, making bones fragile and

susceptible to fractures. This disease usually remains asymptomatic until a fragility fracture

occurs; patients who sustain one fracture are at a much higher risk of additional fractures

(Klotzbuecher et al., 2000). Osteoporotic fractures are associated with high human and health

care costs including loss of mobility, decreased independence, high financial cost and, in the case

of hip fractures, death (Wiktorowicz et al., 2001; Goeree & O'Brien B, 1996).

Effective treatment options and clinical practice guidelines for OP care and prevention of future

fractures exist (Brown & Josse, 2002; Brown et al., 2006) however, less than 32% of patients

who sustain a fragility fracture receive appropriate OP management (Elliot-Gibson, Bogoch,

Jamal, & Beaton, 2004). Barriers to more effective OP management are well documented and

associated with both the patient and the healthcare environment. Patient-related barriers include a

lack of knowledge of OP and consequences of fractures, as well as the perception that treatment

is ineffective and that fracture healing is the end of the health concern (Kaufman et al., 2003;

Meadows et al., 2005). It is therefore not surprising that many patients are reluctant to undergo

further testing and treatment and that medication initiation and adherence remain suboptimal

(Gold, Alexander, & Ettinger, 2006; Elliot-Gibson, Bogoch, Jamal, & Beaton, 2004). Medication

initiation and adherence are ongoing health behaviours that require patients to make and sustain

changes to their existing behaviour. Whether patients will change their behaviour and

initiate/adhere to the prescribed therapy is, in part, dependent on the patients’ perception of the

state of their health (Gold et al., 2006; Wichowski & Kubsch, 1997). Patients’ perception of their

health plays a central role in individual-level theories of health behaviour change. Major theories

of health behaviour change such as the Health Belief Model (Rosenstock, Strecher, & Becker,

1988), Teachable Moment (McBride, Emmons, & Lipkus, 2003), Transtheoretical Model

(Prochaska & Velicer, 1997), Theory of Planned Behaviour (Fishbein M & Ajzen I, 1975) and

11

Social Cognitive Theory (Bandura, 2004) have been developed to explain how and why

individuals change their behaviour. Each of these theories is characterized by a defining

component. Patients’ perception of their susceptibility to a condition is a defining component of

the Health Belief Model; self-efficacy (one’s sense of control and confidence in ability to take

action) is a core feature of the Social Cognitive Theory; the Teachable Moment theory

emphasizes the importance of the affective reaction to a cueing event (event that triggers action);

behavioural intention is a key factor of the Theory of Planned Behaviour; and the

Transtheoretical Model is characterized by the idea that change is staged and not immediate.

Although these theories propose different models of behaviour change, they also share many of

the same features: in order for health behaviour change to occur, most of them would postulate

that patients need to first perceive a health threat, see themselves as susceptible to it, believe that

benefits of the behaviour change outweigh the barriers, perceive themselves as having enough

ability to perform this change and perceive the behaviour as socially desirable. Key concepts of

major theories of behaviour change and their definitions are presented in Table 2.1 (on the next

page).

Theories of health behaviour change have been applied to a variety of behaviours such as

smoking cessation, dietary change, weight control, exercise acquisition, HIV prevention and

chronic disease prevention and management (Becker et al., 1978; Glanz et al., 2002; Munro,

Lewin, Swart, & Volmink, 2007; Prochaska et al., 1994). Theories of health behaviour change

have also been used in the context of OP, mainly in the area of primary prevention (Hsieh,

Wang, McCubbin, Zhang, & Inouye, 2008; Popa, 2005; Wallace, 2002). Some theories have also

been applied to the context of OP management and medication initiation: the Health Belief

Model was used to examine the role of OP beliefs and attitudes in antiresorptive medication use

in community-dwelling women (Cline, Farley, Hansen, & Schommer, 2005) while Gibson’s

patient-centered, self-empowerment model was used to investigate the potential of an

Osteoporotic Pain Program (Jensen & Harder, 2004). Mauck and colleagues utilized a stage of

change model (Weinstein Precaution Adoption Process Model) to examine patients’ decision to

accept OP treatment after hip fracture (Mauck et al., 2002). The same theoretical framework was

used to explore barriers to OP care from the perspective of female fragility fracture patients

(Edwards, Iris, Ferkel, & Feinglass, 2006). Gold and colleagues implemented a self-management

12

OP program for post-menopausal women with OP (or OP risk) that incorporated elements of the

Social Cognitive Theory (Gold & Silverman, 2004).

Table 2.1 Main concepts of five major individual-level theories of health behaviour change Moderating factor = + Main factor = + + Defining factor = defining

Concept

Definition

Health Belief Model

Social cognitive theory

Teachable moment

Theory of planned behaviour

TTM Stages of change

Predisposing factors

Pre-existing factors that may influence the behaviour

+ + +

Cueing event Event that triggers ‘readiness’ to change or provides a cue to action

+ (cues to action)

defining

Affective response

Emotional reaction to a cueing event + +

Perceived susceptibility

to a condition

One’s perception of threat risk or belief about the likelihood of developing a condition

defining + +

Perceived severity or seriousness

One's belief about seriousness of a condition/threat and its consequences

+ +

Perceived benefits

One’s beliefs about the positive outcomes associated with a behaviour

+ + + +

+ + + + +

Perceived Barriers (to

taking action)

One’s perception of the obstacles to specific behaviour/action

+ + + + + + +

Self-Efficacy One’s sense of control and confidence in ability to take action

+ defining + + + +

Social approval

/norm

One’s perception about the social context and the extent to which the behaviour is socially desirable

+ + + + + +

Proximal goals

(intention)

Behavioural plans predictive of performing behaviour

+ + defining

Staged change

Behaviour change involves movement through a sequence of discrete stages

defining

13

A wide range of post-fracture interventions have been created and implemented within the

orthopaedic environment as this is a setting where the concentration of fragility fracture patients

(and cost-effectiveness of interventions) is the highest. These interventions range from sending

reminders and education to GP’s, to complex programs where a dedicated screening coordinator

identifies, educates and coordinates testing and treatment of possible OP (Elliot-Gibson, Bogoch,

Jamal, & Beaton, 2004). It is not known whether behaviour change theories have been used to

create any post-fracture OP interventions within the orthopaedic environment.

Given that post-fracture OP management is a multi-stage process requiring patients to make and

sustain changes to their existing behaviour, it is important to evaluate whether models of

behaviour change have been used in this context. The purpose of this literature review is to

determine whether, and to what extent, current post-fracture interventions utilize theories of

behaviour change and whether those that are theory-based are more successful in changing

health behaviours.

2.2 Methods This literature review was based on studies identified as part of a broader systematic review of

non-surgical OP interventions in the orthopaedic environment (Sale J, under review). The initial

search using keywords “osteoporosis” and “intervention” was conducted in CINAHL, EMBASE

and OVID Medline databases in May 2008; an additional search was conducted in September

2008. Articles were limited to those published in English in the last 20 years that contained the

above keywords in either the title or the abstract (n=608). These articles were further limited to

333 articles that focused on post-fracture OP care in the orthopaedic environment. Letters,

commentaries, literature reviews and articles one page in length or less were discarded. The 333

articles were then restricted to patient-targeted, primary studies that aimed at improving post-

fracture OP care. There were 35 articles that fulfilled these criteria based on a full article review

(Figure 1.1). The 35 studies included in the final review were described in terms of study design,

target population, interventions and outcomes and examined for explicit mention of an

underlying theory.

14

608 articles identified through title

and abstract screen Keywords: OP and non-

surgical intervention

333 articles identified through title

and abstract (full article if needed)

35 articles identified

through full article review

Restricted to articles on post-fracture care. Excluded letters, commentaries, literature

reviews, articles that were 1 page in length.

Restricted to: - primary studies (no literature reviews) - post-fracture population - patient-targeted interventions - studies aimed at improving OP care

Figure 2.1 Search strategy for the literature

review of post-fracture osteoporosis interventions

If the authors did not explicitly state that they used a theory, we looked for a description of a

theoretical framework consistent with one of the major behaviour change theories. To do this, we

searched for presence of constructs from theories of behaviour change and examined whether

they were integrated in a way suggested by one of the theories of behaviour change. Appendix I

table provides a description of all the studies in terms of study design, target population,

theoretical framework, interventions and outcomes. Eight studies contained descriptions of three

programs; those that described the same program were grouped together.

15

2.3 Results Thirty-five studies were included in the final review (Ashe et al., 2004; Astrand, Thorngren, &

Tagil, 2006; Astrand, Thorngren, Tagil, & Akesson, 2008; Bessette et al., 2008; Bliuc et al.,

2006; Blonk, Erdtsieck, Wernekinck, & Schoon, 2007; Bogoch et al., 2006; Che, Ettinger, Liang,

Pressman, & Johnston, 2006; Chevalley et al., 2002; Cuddihy et al., 2004; Davis, Guy, Ashe,

Liu-Ambrose, & Khan, 2007; Edwards et al., 2005; Feldstein et al., 2006; Feldstein et al., 2007;

Fraser & McLellan, 2004; Gallacher, 2005; Gardner et al., 2005; Harrington, Barash, Day, &

Lease, 2005; Harrington & Lease, 2007; Hawker, Ridout, Ricupero, Jaglal, & Bogoch, 2003; Ho

et al., 2006; Johnson, Petkov, Williams, Via, & Adler, 2005; Kuo et al., 2007; Langridge et al.,

2007; Laslett et al., 2007; Levasseur et al., 2007; Majumdar et al., 2004; Majumdar et al., 2007;

Majumdar et al., 2008; McLellan, Gallacher, Fraser, & McQuillian, 2003; Quintos-Macasa et al.,

2007; Schmid, Henzen, Schlumpf, & Babst, 2004; Skedros, 2004; Tosi, Gliklich, Kannan, &

Koval, 2008; Vaile, Sullivan, Bennett, & Bleasel, 2007). Nineteen of these studies described OP

screening or management programs incorporated in clinical care, while 16 studies described

study-specific interventions: randomized controlled trials (n=7); controlled trials (n=2); and other

interventions (n=7). One study (Bessette et al., 2008) was a randomized controlled trial that was

also a part of a provincial management program.

We identified seven main elements that were present in patient-targeted interventions: provision

of patient-specific or general OP information (written or verbal), OP management-related

consultation, medical assessment (by one or more health care professionals), questionnaires

assessing patient OP self-reported risk factors, knowledge and/or attitudes, recommendations

(and/or a letter) to see a primary physician or another health care professional, BMD booking (or

asking a patient to arrange his or her BMD) /follow up with a primary physician/specialist, and

treatment recommendation/prescription. Written information (only) was given to patients in five

interventions. Of these, three letters were personalized, informing the patient about his/her own

risk for OP/future fractures. All other interventions included some degree of verbal information.

Twenty interventions involved a consultation (in person or over the phone), which consisted of

discussing OP-related issues such as risk factors, the link between the fragility fracture and OP,

medication and supplements, lifestyle modification (diet, exercise, fall prevention),

communication with a primary physician and BMD results. Three studies relied only on

orthopaedic surgeons to give patients a few key messages. Patients were recommended to see

16

their primary physician in 15 interventions: in seven of those patients were given a letter to take

to their physician. Medication was offered or prescribed in six interventions. Physicians were

directly contacted in 20 cases. Ten interventions were directed at patients only (physician was

not contacted or it was left up to the patient to contact physician or arrange testing).

None of the studies specified an underlying theoretical base for the interventions. Descriptions of

the interventions themselves also did not reveal an underlying theoretical framework. This lack

of theoretically-based studies prevented us from making a comparison between theoretically and

non-theoretically based interventions.

Apart from examining the studies for explicit or implicit use of theory, we also looked for the

presence of any constructs from theories of behaviour change and found three studies which

mentioned constructs that were, what we felt could be, elements of a framework. The first study

that included a construct suggestive of a framework was the study by Vaile and colleagues (Vaile

et al., 2007), which described an intervention aimed at fragility fracture patients with “fresh

fracture pain.” Aiming a post-fracture intervention at such patients fits the Teachable Moment

framework because this framework emphasizes the importance of using a strong emotional

response to a cueing event in building a “teachable moment” and motivating subsequent

behaviour change. The concept of the “teachable moment” was previously mentioned in the

context of post-fracture care by two authors (Boden, Einhorn, Morgan, Tosi, & Weinstein, 2005;

Meadows et al., 2005): one suggested that fractures were excellent teachable moments to educate

patients about the importance of OP care while the other concluded that many patients have

misperceptions about their OP and fracture risks and that physicians can use fractures as

teachable moments to deliver health information and, more importantly, make this information

relevant for patients.

The second study that made use of a construct consistent with a theoretical framework was a

study that tested a Patient Empowerment and Physician Alert intervention (Davis, Ashe, Guy, &

Khan, 2006). This intervention targeted patients’ awareness of OP and encouraged them to play a

more active role in the management of this disease. To participate in the management of one’s

own disease, patients need to feel confident in their ability to do so. Confidence in one’s ability

to take action, also known as self-efficacy, is a core feature of Bandura’s Social Cognitive

Theory. This theory has been the basis for many chronic disease self-management programs,

17

which seek to empower patients and help them take control of their own health (Gold &

Silverman, 2004; Lorig et al., 1999).

Similar to the study that incorporated patient empowerment, the intervention used by Astrand

and colleagues (Astrand et al., 2006; Astrand et al., 2008) encouraged patients to take on a more

active role in the management of their health. The premise of this intervention was that the

patients could be trusted to seek appropriate medical advice and make relevant decisions. This

premise assumes that patients will make rational decisions based on the information provided

and that the behaviour in question is completely voluntary. Such rational appraisal of benefits

and barriers to action is a characteristic of the Health Belief Model. The Health Belief Model,

however, also places an emphasis on patients’ perception of their susceptibility to a disease while

the premise of “trusting patients to make relevant decisions” assumes that patients will perceive

the health threat, and the information given, in the way providers of the information intended it

to be perceived (as meaningful and relevant).

The 35 studies included in this review utilized a variety of post fracture interventions; however,

as previously mentioned, none of them reported using an underlying theoretical base, therefore

we were unable to make a comparison of outcomes between theoretically and non-theoretically

based studies. The lack of theoretically-based studies, however, points to a currently under-

utilized area of behaviour change research that could be applied to post-fracture interventions in

order to make them more effective.

2.4 Discussion The findings of this literature review indicate that despite a variety of post-fracture interventions

and literature supporting theories of behaviour change, current patient-targeted post-fracture OP

interventions do not report using theoretical frameworks.

Formal application of theories of behaviour change would improve our understanding of the

mechanisms involved in behaviour change and could contribute to more effective post-fracture

interventions. It is known that patients’ perception of their need for OP treatment can be a barrier

to effectively managing OP (Kaufman et al., 2003). Creating an intervention without considering

18

the fact that patients may not perceive the need for treatment in the same way as health care

practitioners, may predispose this intervention to failure. Patients are not likely to initiate or

adhere to treatment if they do not believe that they have a health problem that is serious and

which needs to be managed in a way prescribed by health care practitioners. Using frameworks

that account for patients’ perceptions could prove very useful in this case. Theories of behaviour

change would suggest that the interventions where the patient is either not engaged or is a

passive recipient of information would not be very effective in producing the desired behaviour

change, particularly in the long run. An intervention that would have the potential to generate

desired behaviour changes involves personalized consultation that addresses OP management

issues. In order to be effective, this intervention would need to be based on the individual

patient’s perception of factors such as susceptibility, self-efficacy, subjective norm, barriers to

and benefits of behaviour change.

Each of the five major theories of health behaviour change can be applied to the context of OP to

explain how and why patients do or do not initiate OP management including a follow up with a

physician, further testing, initiation or staying adherent to a prescribed treatment. Health Belief

Model, for example, would suggest that patients’ likelihood of behaviour change, such as

treatment initiation, depends on the following factors: whether patients perceive OP as a serious

threat, themselves as susceptible to it and treatment as effective and doable; whether they believe

they are able to adopt the new behaviour and whether there is a cue to action stimulus that would

motivate the change. Application of the Health Belief Model to the context of OP treatment

initiation would therefore include personalizing risk factors, explaining consequences of

untreated OP, discussing benefits of treatment and ways to overcome barriers, as well as

providing reminders, training and guidance on how to successfully initiate OP treatment.

Theory of Planned Behaviour, on the other hand, proposes that the most immediate determinant

of peoples’ behaviour is their intention to perform that behaviour. Since the theory assumes that

intention is determined by attitudes, subjective norm and perceived behavioural control, the

intervention would start with assessing the patient’s attitude (towards the specific behaviour,

such as treatment initiation), the patient’s perception of the extent to which this behaviour is

under his or her control and the patient’s belief that significant others would support it. An

effective intervention would then be tailored to the patient’s intention to initiate treatment.

19

Social Cognitive Theory, well-known for introducing the concept of self-efficacy (belief in one’s

capability to actually produce a particular behaviour), would suggest that successful behaviour

change depends on whether patients feel that they have enough self-efficacy to adopt a new

behaviour. Self-efficacy beliefs do not operate in isolation; they interact with outcome

expectations, perceived environmental barriers, and facilitators and goals, all of which should be

taken into consideration. The intervention based on Social Cognitive Theory would therefore

consist of increasing the patient’s (perceived) self-efficacy for treatment initiation by setting

achievable goals, providing ways for the patient to learn the necessary skills and continually

discussing progress.

Transtheoretical Model or Stages of Change model, which asserts that changes in patients’

behaviour are made gradually and in distinct phases, would suggest that OP interventions are

only going to be effective if adapted to the stage of change in which the patient is currently.

According to this theory, an individual begins at a pre-contemplation stage, where the change is

not even considered, and progresses through to the maintenance stage, where the behaviour has

been maintained for at least six months.

Teachable Moment as adapted by McBride and colleagues (McBride et al., 2003) builds on many

of the previously-mentioned models and focuses on the cueing event that, if strong enough,

motivates behaviour change. In order for an event to be considered a Teachable Moment event,

underlying constructs need to be present, including an increase in perceived personal risk, a

strong emotional response and a re-definition of self concept or social role. In a 1999 study on

alcohol-related emergency room admittances, brief motivational interviews during “teachable

moments” were shown to reduce future alcohol-related consequences (Monti et al., 1999). The

same concept could be applied to motivating fragility fracture patients to seek OP testing and

care at the time of fracture. A fracture can be used as a cueing event to create a teachable

moment and motivate patients to adopt a new health behaviour, such as OP treatment.

One could argue that individual-level theories of health behaviour change do not adequately

account for external factors which play a big role in the context of OP management. The external

factors, such as access to bone mineral density testing, health-practitioner’s lack of knowledge of

current OP guidelines and lack of time to address OP concerns, are all associated with the system

and the health care practitioner. With the exception of Social Cognitive Theory, individual-level

20

theories of health behaviour change do not explicitly include these factors in the equation.

Instead, they suggest that external factors operate through other concepts included in the theory

(Glanz et al., 2002). For example, the concept of barriers and facilitators of OP management

incorporates external factors as perceived by an individual patient. Even though the importance

of external factors is undeniable, one could argue that effective OP management ultimately

depends on the individual patient’s choices. Gold and Silverman suggested that improvements in

chronic disease management can only occur if individual patients take responsibility for their

health behaviour (Gold & Silverman, 2004) while Ryan and colleagues pointed out that, outside

of acute care and despite technological advances, the effectiveness of many health care

interventions depends on the patient’s behaviour (Ryan, Patrick, Deci, & Williams, 2008).

Our review has limitations. Only articles in English found in CINAHL, EMBASE and OVID

Medline databases were included. Unpublished, non-peer reviewed work and abstracts from

conferences were not included in the review, which could have resulted in the omission of some

interventions that were based on theories of health behaviour change. A possibility that not all

patients will benefit from the same theoretical approach should be considered as a limitation of

applying behaviour change theories to the post-fracture context. In our previous focus group

work, we came across two groups of patients: those who were taking steps to actively manage a

health issue; and those who entirely depended on their physician for all health-care needs

(Beaton et al., 2007). It is therefore possible that the approach which puts emphasis on the

patient may not work in the latter category or in cases of older patients with dementia.

Future post-fracture interventions should be built on previous research in the area of OP

management. Theoretical frameworks previously suggested in the context of OP management

(Boden et al., 2005; Meadows et al., 2005) should be tested. More research identifying

intervention gaps from the patient’s perspective should also be conducted. Previous research, for

example, found that many fragility fracture patients do not even perceive a link between their

fragility fracture and OP (Giangregorio et al., 2008). As a result, these patients do not perceive

themselves to be susceptible to OP and/or future fractures. Identifying patients who are more or

less likely to make the link between OP and their fracture could help us design more effective

post-fracture interventions tailored to different groups of patients.

21

For chronic disease interventions to be successful, patients need to initiate and maintain new

health behaviour(s). Health behaviour change theories specify processes that regulate behaviours

and account for patient perception of health which has been shown to be an important factor in

the patient’s decision to initiate and adhere to the prescribed care (Wichowski & Kubsch, 1997).

Future research should explore the application of theories of health behaviour change to post-

fracture interventions. Such interventions would contain a health message that is tailored to the

specific patient and his or her OP knowledge and beliefs. In light of the fact that OP is under-

diagnosed and under-treated, and that fragility fracture patients do not perceive themselves to be

at risk for OP, it is important to design more effective post-fracture interventions with a focus on

patients’ perception of their health.

A modified version of this chapter has been accepted for publication. The final publication will

be available at www.springerlink.com

22

Chapter 3 Critical Appraisal of the Ontario Osteoporosis Strategy’s Fracture Clinic Screening Program Database as a Source of Data for the

Thesis

3.1 Introduction The first objective of this thesis, a review of current OP post-fracture interventions in the context

of theories of health behaviour change, has been described in Chapter 2. The review showed that

theories of behaviour change have not been integrated into post-fracture interventions and that

more research into the role of patient perceptions is necessary in this area.

The second objective of this thesis is to further examine patient perceptions by analysing whether

patients perceive the link between their fracture and OP. The analysis of the OP-fracture link will

be described in Chapter 4. This chapter will provide background information to set the stage for

the OP-fracture link analysis by describing and critiquing the source of data.

3.2 Background and description of the database Data used in the OP-fracture link analysis were derived from the Fracture Clinic Screening

Program database. The Fracture Clinic Screening Program, one of five components that comprise

the Ontario Osteoporosis Strategy, was implemented by Osteoporosis Canada, under the

directive of the Ministry of Health and Long Term Care (MOHLTC) of Ontario (Jaglal et al.,

2010). As part of this program, Osteoporosis Canada placed screening coordinators in designated

fracture clinics across Ontario with the long term goal of increasing secondary prevention among

fragility fracture patients and decreasing the rates and associated costs of subsequent fractures.

The clinics were identified by estimating yearly volumes of fragility fracture patients using

administrative data. Sixty clinics with high or moderate volumes were identified as sites where it

could be cost effective to place a screening coordinator to perform screening and facilitate follow

up for OP management. At the time data used in this thesis were collected, there were 19

coordinators working in 36 (of 60) clinics across Ontario.

23

As part of the Fracture Clinic Screening Program, the role of the screening coordinators was to

identify, assess, refer and provide education to fragility fracture patients who presented

themselves at the clinics where the coordinators were located. The process started by

coordinators identifying patients 50 years of age and older who had sustained a possible fragility

fracture (i.e., who fractured due to low trauma, defined as a fall from standing height or less) of

the wrist, clavicle, elbow, spine, pelvis, femur, tibia/fibula, ankle or other selected sites.

Screening coordinators first approached the identified patients in order to confirm that they had

sustained a low trauma fracture. Patients who sustained a low trauma fracture (or were referred

by a surgeon) were interviewed by screening coordinators in order to complete the baseline

questionnaire. If the patient or his/her proxy was unable to answer the questions and was

therefore unable to complete the baseline questionnaire, the coordinators would complete only

the first section of the questionnaire which included information on patient’s age, primary

language, first three digits of the postal code (FSA), fracture (site, etiology, date) and reasons for

not completing the questionnaire (language/mental/physical barriers or refusal). Other sections of

the questionnaire covered OP risk factors, OP diagnosis and treatment history, bone mineral

density (BMD) testing, demographics, documentation of intervention and patient’s OP

knowledge and OP beliefs. With the exception of patient’s OP knowledge and beliefs, the

questionnaire could be filled out by an accompanying person (proxy), if the patient was unable to

do so.

The intervention that patients received as part of the Fracture Clinic Screening Program

consisted of: verbal recommendations for further assessment of bone health and BMD testing,

recommendations for Calcium and Vitamin D intake and provision of (written) general OP

educational materials, Program Information Letter and other Osteoporosis Canada resources

(details are available in Appendix II). Screening coordinators would also send a letter to the

patient’s family physician with a request for further assessment of bone health (unless the patient

requested that his or her family physician not be contacted).

All patients who were not diagnosed and/or treated for OP at the time of screening were asked if

they would agree to a telephone follow up at three and/or six months. Those who agreed were

called at the appropriate time and asked about post-fracture follow up with their family physician

and new testing, diagnosis and treatment information. The same OP knowledge/beliefs questions

24

asked at baseline were also asked at follow up. Patient flow through the Fracture Clinic

Screening Program is depicted in Figure 3.1.

Patients ≥ 50 years of age, potential low trauma fracture

approached by coordinators to establish fracture etiology

Low trauma patients (or those referred by a surgeon)

completed baseline questionnaire

Undiagnosed and/or untreated patients were asked to agree to follow

up

Screening stopped for moderate or high trauma, missed patients or those who refused to participate in the program

Interaction stopped for patients who were previously diagnosed or treated for OP

Interaction stopped for patients who were lost to follow up

Figure 3.1 Patient flow through the

Fracture Clinic Screening Program

Patients followed up at three or six months to complete follow up

questionnaire

Data collected at baseline and follow up were used for ongoing quality improvement within the

program and for reporting to the MOHLTC. A subset of data was also used for research: all

patients (or their proxies) capable1 of providing their consent were asked whether they would be

1 As per coordinator’s assessment

25

willing to: (a) allow the information gathered as part of the program to be used for research

purposes; and (b) provide necessary personal health information to allow linkage to the

MOHTLC databases for research into future fracture rates and OP-related care. Patients’ consent

was documented in the database and only those patients who consented to research use of their

baseline and follow up information were included in the research dataset. Approval of the site’s

research ethics board was also obtained prior to consenting patients for inclusion in the database.

3.3 Critique of the database: establishing its suitability to the research question

In this section, I will assess the strengths and weaknesses of the Fracture Clinic Screening

Program as a source of data for my research question. To make this assessment, I will use

elements from Connell’s (1987) framework for identifying and dealing with issues inherent in

working with large health care research databases. To understand possible data flaws, Connell

suggested examining the following categories: data collection methods (population, sampling

strategy, time frame, how were data collected, inclusion criteria, comparison of subgroups),

database structure (unit of observation, content of each observation, duplicates), quality of data

elements (coding, missing observations, out of range values) and linkage of datasets. As part of

the same framework, Connell also listed specific strategies such as manual examination of

records for data flaws, frequency analysis of all variables and inspection for unusual

distributions, out-of-range and missing values (Connell, Diehr, & Hart, 1987).

My description and critique of the Fracture Clinic Screening Program database will be based on

Connell’s framework and organized as follows:

1. Purpose and content of the database, sampling strategy, comparison of the sample to the target

population and data collection methods

In addition to discussing the original purpose and content of the database, I will examine the

sampling strategy (including time-frame and the inclusion/exclusion criteria), comparison of

the sample to the target population, data collection methods and possible biases associated

with data collection.

26

2. Database structure, content of each observation, duplicates and coding

In this section, I will review the overall structure of the database, the content of each

observation and the data quality issues (possible duplicates and coding of variables).

3. Manual inspection of data for data flaws and an assessment of the overall data quality

To assess features that may indicate a pattern of errors, I will perform a manual inspection of

three samples of raw data using the first, last and random 10% observations. I will critique

any patterns of miscoding, missing observations or out of range values that may be found.

Overall quality of the data will be assessed as well.

3.3.1 Purpose and content of the database, sampling, comparison of the

sample to the target population and data collection methods

Purpose: One of the first steps in establishing appropriateness of the database to the research

question is to consider the original purpose of the database and its content (Connell et al., 1987).

The Fracture Clinic Screening Program database was created as part of a province-wide

screening program whose short-term objectives were to enhance patient OP knowledge, the

awareness of the OP-fracture link and the need for testing, as well as to increase OP awareness of

healthcare providers and increase referrals of fragility fracture patients to family physicians for

assessment and treatment. The database was designed to collect data for ongoing evaluation of

the screening program (quality assurance) and future research into post-fracture OP care and care

gaps in Ontario, with the final goal of improving OP care in the province. Variable selection was

done by the research team and Osteoporosis Canada, based on the available literature and input

from experts in the field. Specific variables were selected in order to address a number of

possible research questions and to inform quality assurance initiatives. Variables were taken

from both previously validated instruments and previous studies to allow comparability with the

literature. Even though the research question asked in this thesis was not established prior to data

collection, the database contained variables (predictors and the outcome of the OP-fracture link)

that could be applied to the research question being addressed in this thesis. In addition, the

database contained the data from fragility fracture patients, the target population of this thesis.

27

Given that the database variables and the population met the needs of this research (to inform

efforts to optimizing post-fracture OP care in an orthopaedic setting), it can be concluded that the

Fracture Clinic Screening Program database was a good source of information for the research

question to be answered.

Content: At the time data used in this research were collected, the Fracture Clinic Screening

Program database consisted of self-reported baseline and follow-up data from fragility fracture

patients who were screened by 19 screening coordinators at 36 sites across the province.

Fragility fracture patients who were over the age of 50 at the time of the fracture were the target

population for the screening program and its research component. Baseline information was

collected on all patients who fit the screening criteria (over the age of 50 with a low trauma

fracture or surgeon-referred) while the follow up was done only with those patients who were not

diagnosed and/or treated for OP at baseline. Only these undiagnosed and untreated patients were

followed up because (a) the resources were not sufficient to follow up with everyone and (b) the

undiagnosed and untreated patients were a priority for the intervention (assessment, education,

referral) as those previously diagnosed and/or treated for OP had already taken steps towards OP

management. Patient data collected at baseline included questions on OP risk factors, OP

diagnosis and treatment history, BMD testing, demographics, documentation of

intervention/patient consent and patient’s OP knowledge and beliefs. The follow up

questionnaire also included questions about OP knowledge and beliefs in addition to follow-up

questions about post-fracture visits with a family physician and testing, diagnosis and treatment

information. Questions asking about OP knowledge and beliefs were derived from previous

research. OP knowledge was measured by four items from the “OP and You” knowledge

questionnaire (Brenneman, Blau, Chen, & Abbott, 2002) previously selected and validated by

Cadarette (Cadarette, Gignac, Beaton, Jaglal, & Hawker, 2007). Questions on OP beliefs

included questions on patients’ perception of OP treatment, the link between their fracture and

OP, the quality of patients’ bones as well as patients’ perception of benefits and barriers to OP

pharmacotherapy. Questions on both perceived benefits and barriers to pharmacotherapy were

derived from previous research (Cadarette, Beaton, & Hawker, 2004) and based on the

“Osteoporosis Health Belief Scale” (Kim, Horan, Gendler, & Patel, 1991). Perception of benefits

of OP pharmacotherapy was measured by a 5-item OP benefits scale while barriers were

28

assessed by two individual items. The measures are described in greater detail in Chapter 4 with

additional information presented in Appendix III (construct list).

Even though the database contained the variables of interest to this thesis, there are other

concepts which could have also been explored. For example, Tugwell’s concept of the patient as

an “effective consumer” (Tugwell, Santesso, O'Connor, & Wilson, 2007) or a related concept of

patient’s health literacy (McCabe, 2006), would have also been of interest but were not explored

because they were not captured by the database. The concept of health literacy, defined as ”the

degree to which individuals have the capacity to obtain, process and understand basic health

information and services needed to make appropriate health decisions" (National Library of

Medicine, 2010) could be a potential predictor of the OP-fracture link. Patients who have high

health literacy may be more likely to make the OP-fracture link because they would also be more

likely to obtain relevant health information and apply it to themselves appropriately. Whether

this ability to obtain, interpret and apply relevant health information is predictive of making the

OP-fracture link at follow up was left unexplored. Other concepts that could be useful predictors

of the OP-fracture link included cultural and social issues, education level and co-morbidities.

Despite the database not capturing every variable that would be of interest for this thesis’

research question, the database contained most variables of interest making it a rich and fitting

source of data.

Sampling strategy: The sample used to answer the research question on the OP-fracture link

was derived from the Fracture Clinic Screening Program research database. At the time of

analysis, this research database contained data from consenting patients who were screened

between December 2007 (when the consenting process started) and September 2009 (as per the

REB approval at the time2). For purposes of the OP-fracture link analysis, the sample was

further defined to include patients who:

a) Were eligible for follow up (not diagnosed or treated for OP at baseline) and agreed to

follow up

2This REB approval allowed for research use of baseline data collected up until September 30, 2009 and the follow up data of the same patients obtained thereafter.

29

b) Were followed up by December 31, 20093 and had complete data for the OP-fracture link

question at both baseline and follow up

c) Did not make the link between OP and their fragility fracture at baseline

In using the above inclusion criteria, the research dataset (containing only data from consenting

patients) was reduced from 4,657 fragility fracture patients over the age of 50 who were eligible

for follow-up to 853 patients who fulfilled the rest of the inclusion criteria (Figure 3.2 Sampling

strategy for the OP-fracture link analysis ).

Figure 3.2

Sampling strategy for the OP-fracture link analysis

n=4657 Eligible for follow up

3 December 31, 2009 was chosen as a cut-off date because the OP-fracture link analysis started in January 2010.

n=4011 Did not refuse follow up

n=2207 Followed up with

n=916 Had complete data on OP-fracture

link at baseline and follow up

n=853 did not make the OP-fracture link

at baseline

30

The inclusion criteria reduced the sample size but also helped meet the requirements of the OP-

fracture link research question. These inclusion criteria were set for the following reasons:

a) The research question examined the change in the OP-fracture link over time; therefore

those ineligible to be followed up (treated and/or diagnosed with OP) were excluded as

they would not have the necessary follow up data. Second, including patients who were

previously treated and/or diagnosed with OP would also make the direction of the OP-

fracture association unclear as previously diagnosed and/or treated patients have already

taken steps towards OP management and could make the OP-fracture link at follow up as

a result of being diagnosed and/or treated at baseline. The OP-fracture link analysis

focused on identifying clinically-useful predictors of making the link which was

hypothesized to lead to taking steps towards OP management: a finding that patients who

were already diagnosed and/or treated at baseline make the OP-fracture link at follow up

would therefore have not been clinically useful.

b) Since the research question examined predictors of the OP-fracture link at follow up, it

was necessary to know which patients did or did not make the link at baseline and follow

up. Only patients who had data on this variable (OP-fracture link) at both baseline and

follow up were included in the analysis.

c) Because the research question aimed to identify predictors of the OP-link, including

patients who had already made the link would not contribute to the findings. The analysis

was therefore further restricted to those who did not make the OP-fracture link.

Comparison of the sample to the target population: Whenever conclusions are drawn from

sample-specific observations, there is a possibility that these conclusions may not be

generalizable because the sample may be systematically different from the population from

which it was drawn (Fletcher RH, 2005). Such sampling bias can occur anytime a sample is

drawn from a population of interest. To examine whether the sampling bias was present in this

study, I compared my study sample to the target population. Since I was only interested in

generalizing the findings to patients who did not make the OP-fracture link at baseline, I

compared my sample (n=853) to patients who were eligible for follow up, who did not make the

OP-fracture link at baseline (n=3090) and who were not in my study sample (3090 – 853 =

2237). Figure 3.3 shows the comparison that took place.

31

Figure 3.3 Comparison of the sample (n=853) to the

target population (n=2237)

n=4011 Did not refuse follow up

n=2207 Followed up with

n=916 Had complete data on OP-

fracture link

n=853 did not make the OP-fracture

link at baseline

n=4657 Eligible for follow up

n=3090 did not make the OP-

fracture link at baseline

Compare n=2237 (3090-853=2237)

to the sample n=853

The results of the comparison are shown in Table 3.1. The sample and the rest of the population

were compared on possible confounders, predictors and the outcome of interest. They differed on

four variables: work status, perception of OP treatment benefits, steroid use and one OP

knowledge item “If a person has OP, something as simple as lifting a bag of groceries can break

a broke”. Specifically, the sample used in the OP-fracture link analysis consisted of fewer

patients who were employed (30.1% vs. 35.4%), more oral steroid users (6.3% vs. 4.3%), more

patients who perceived higher benefits of OP drug treatment (4.2/5 and 4/5) and more patients

who correctly answered the above-mentioned knowledge item (84% vs. 78%). Fewer employed

patients in the sample could have been a result of the follow up procedure that coordinators used

where most follow up calls were made during working hours (9am-5pm), resulting in more

complete data among those who were not working and were available during these hours. The

second variable that my sample and the comparison group differed on was the perception of OP

drug treatment benefits. Even though the difference was statistically different in two groups, both

groups perceived four out of five benefits to OP pharmacotherapy.

32

Table 3.1 Description and comparison of the sample n=853 to the target population n=2237 Variables n=853 n=2237 p value

Mean age (SD) range 66.6 (10.5) 50-94

66.6 (10.8) 50-95 0.97

Gender – Female

79.7%

76.8%

0.08

Marital status: Married/common law

59.7%

62.5%

0.34

Living situation: Live alone

28.7%

26.8%

0.31

Work full or part time

30.1%

35.4%

0.0054

Hip or shoulder fracture (may include others)

25.6%

24.5%

0.56

Previous fracture >age 40

25.2%

25%

0.94

Maternal fracture >age 40

28%

25%

0.09

Previous fall in the last year

26.6%

27.2%

0.73

Feeling unsteady

27%

30.2%

0.08

Current or past steroid use

6.33%

4.31%

0.02

Perceived benefits of OP drug treatment (1-5)

4.2

4

0.0032

% Who perceived bones as thin

10.6%

10%

0.11

Knowledge item: There is a way to prevent OP (reversed)

64.2%

67.7%

0.07

Knowledge item: Bones can be rebuilt once they are thin from OP (reversed)

57.1%

55.5%

0.41

Knowledge item: If a person has OP, something as simple as lifting a bag of groceries can break a bone

84.3%

78.7%

0.0006

Knowledge item: Health problems caused by OP can be life-threatening

69.2%

71.1%

0.33

Perceived barriers item: Taking too many medications

22.2%

20.8%

0.38

Perceived barriers item: Stomach problem limits ability to take drug treatment

17.5%

20.3%

0.1

The difference in use of oral steroids and correctly answering the knowledge item “If a person

has OP, something as simple as lifting a bag of groceries can break a bone” was statistically

33

significant and could have biased the sample; however, the differences were not large (2% oral

steroids and 14% knowledge question). The sample and the population did not differ on other

variables of interest: there were no statistically or clinically important differences in age, gender,

marital status, living situation, type of fracture, existence of previous or maternal history of

fractures, previous falls, feeling unsteady, perception of the bone quality, nor remaining three

knowledge items and perceived barriers to OP drug treatment.

The results of the comparison of the sample to the rest of the population revealed that the sample

did not differ from the population on most variables but was slightly different when it came to

the employment rate, steroid use and perceived benefits of OP treatment. Since the differences

were not large, the results of the analysis should be generalizable to the target population as long

as these differences are taken into account.

Data collection methods: Data collection process used to create the Fracture Clinic Screening

Program database was described in the first section of this chapter. Data collection was

electronic: coordinators entered data directly into a centralized database hosted and maintained

by an external vendor. Other modes of data collection were also available in case direct

connection to the internet and server was unavailable: coordinators could collect data on tablets

and upload them at a later time. In case of electronic data collection failure, paper versions of the

questionnaire were available. The main advantage of electronic data collection was that it saved

time by skipping the data entry step (which can also carry a risk of data entry errors).

In comparison to other modes of data collection, such as mailed surveys, the disadvantage of

data entry done by coordinators was the dependence of data capture on the time that coordinators

had available to perform data collection. Time allocated to certain tasks could also be an issue: a

number of patients were not followed up because of the limited number of follow up phone calls

that coordinators were supposed to make4. However, in contrast to mailed surveys where the

completeness of responses and return of the survey are at the sole discretion of the participant,

data collection done by coordinators allowed for more control over data capture.

4 According to the protocol, coordinators were to make three follow up calls at three months and three calls at six months.

34

A number of systematic errors can occur during the process of data collection. Measurement

bias, a systematic error that poses a threat to the validity of results, occurs when the methods of

measurement are different among different groups of patients (Fletcher 2005). Since data

collection for the purposes of the Fracture Clinic Screening Program involved different screening

coordinators collecting data at different sites, measurement bias could have occurred. To

minimize the possibility of such bias occurring, each coordinator was trained in data collection

by the same consultant and the same protocol for data collection and documentation had been

followed at each of the 36 sites. Quality checks were also built into the system to avoid out of

range values or incorrect skip patterns. As part of ongoing quality assurance efforts, coordinators

were regularly given feedback about data they collected. A private website was created in order

to regularly communicate data quality issues with coordinators.

Another potential bias associated with measurement could have occurred in situations when

coordinators needed to use paper copies of the questionnaire because the electronic data capture

was unavailable. In such cases, coordinators were able to omit entire sections of the

questionnaire and a deviation from skip patterns, which are automatic in the electronic version,

was possible. To examine the possibility of such deviation occurring, a manual examination of

raw data (described in the next section) was undertaken.

Systematic errors are a possibility during any data collection procedure, particularly in multi-site

studies. Fracture Clinic Screening Program placed a number of procedures into place in order to

decrease such errors during data collection and ensure that potential measurement bias was

minimized.

3.3.2 Database structure, content of each observation, duplicates and coding

Structure: The Fracture Clinic Screening Program database was designed with fixed variables

with pull-down menus (multiple choice) and “free text” fields when “other” value was chosen.

Appendix III contains the construct list with baseline and follow-up variables contained in the

database. As part of the quality assurance, a number of smaller datasets, corresponding to

different sections of the baseline and follow up questionnaire, were extracted from the

35

centralized database and sent to the evaluation and research team at St. Michael’s Hospital where

the smaller datasets were merged into a single database. Data cleaning was done as part of the

quality assurance. Screening coordinators were regularly given feedback on the quality of the

data they were collecting.

Content of each observation: In the study sample, each observation referred to one patient and

contained data collected at both baseline and follow up (this was not the case in the Fracture

Clinic Screening Program database where every patient did not necessarily have follow up

information). Each observation/record in the database was marked by a unique study ID referring

to one individual patient. The unique study ID was a seven digit number assigned by the

screening coordinator who performed data collection: first two digits corresponded to the pre-

assigned site ID (indicating which hospital the patient was seen at) and the last five were patient-

specific (in chronological order). Another unique study ID was also assigned by the electronic

system, when the observation was entered into the database. Having two separately-assigned

ID’s (assigned by the system and the person performing data collection) was useful in

identification of errors and possible duplicate data entries.

Duplicates: Checking for duplicates in the unique ID was performed as part of the data quality

assurance. The potential for coordinators-assigned ID errors was high given that this ID field was

in “free-text” format where a number of variations of the same ID were possible. For example, a

patient number 36 from site 20, who would have been identified as “20-00036”, could have also

been identified as “20 036”, “20-00000036” or “00036”). A number of variations of the required

seven-digit format were indeed found during the data cleaning procedure (part of quality

assurance). This issue was resolved by creating a new subject ID that consisted of the last three

digits of the coordinator-assigned subject ID (the part that was correctly formatted) combined

with the ID number specific to the site where the patient was screened. Duplicates were checked

for by examining frequency distributions of both the newly-created subject ID and the subject ID

assigned by the system during data entry. In case of doubt, records were manually checked for

possible duplicate content. All duplicates found were excluded from the overall dataset. All of

this was done as part of quality assurance efforts, prior to extracting the study sample from the

database.

36

Coding: Initial coding structure was agreed upon at the time of variable selection for the

Fracture Clinic Screening Program database. Possible inconsistencies in coding in the study

sample were checked by examining the frequency distribution for all variables. All data elements

were appropriately coded: no changes in coding were found and there were no variables that

were not accounted for. Some categorical variables were re-coded after examination. For

example, the codes “thin” and “definitely thin” were re-coded into one category “thin”; “agree”

and “strongly agree” were re-coded as “agree”; and “disagree” and “strongly disagree” were re-

coded as “disagree”. Distributions of continuous variables were also examined and no out-of-

range values were found. During frequency distribution checks, variables that were marked by a

large number of missing observations were flagged for further sensitivity analysis.

3.3. 3 Manual inspection of data for data flaws

In this section, I made comparisons across three data samples to determine any time-dependent

shifts in data quality. Namely, I evaluated data flaws (undocumented codes, unexpected

distributions and missing values) by manually examining three samples of raw data: the first

10% (observations 1-85), last 10% (observations 768-853) and random 10% (observations

numbered 86-767). I also examined the frequency distribution of the key variables (predictors,

moderators, outcomes) for potential data flaws.

Raw data examination showed that there were no undocumented codes. The only unexpected

distribution was found in the last 10% of the data where 16% of answers for the knowledge

question “The health problems caused by OP can be life-threatening” were marked as “refused”

(27/170). The same question had 12% (10/85) of missing values in the random 10% sample.

The only difference in the structure across the three sections was that there were no missing

values in the last 10% of observations, which could be due to OP knowledge/beliefs questions

becoming mandatory later on in the data collection process (which would also explain the high

number of “refused” in the same section). Table 3.2 shows the distribution of missing values for

the three sections.

37

Table 3.2 Manual examination of raw data: missing values in three data samples

Variable First 10%

(n=85)

Random 10%

(n=85)

Last 10%

(n=85) Age 0 0 0

Gender 0 0 0

Marital status 0 0 0

Living situation 0 1 0

Work status 0 0 0

Fracture type 0 0 0

Previous fracture 0 0 0

Maternal fracture 0 1 0

Fall in the last year 0 3 0

Feeling unsteady 0 2 0

Current or past oral steroid use 0 0 0

OP knowledge items • There is a way to prevent OP (rev.) • Bones cannot be rebuilt once they are thin from OP • If a person has OP, something as simple as lifting a bag of

groceries can break a bone • The health problems caused by OP can be life-threatening

0 1 1 0

1 1 0

10

0 0 0 0

Perceived benefits of OP pharmacotherapy • Drug treatments can help build strong bones • You would feel good about taking drug treatments to treat OP • Drug treatments can cut the chances of broken bones • You would consider taking drug treatments to prevent broken

bones • If your doctor advised you to, you would take drug treatments

to prevent broken bones

3 3 3 2 6

1 0 2 0 1

0 0 0 0 0

Perception of bones as thin 0 0 0

Perceived barriers to OP pharmacotherapy • You are taking too many medications • You have a stomach problem that limit your ability to take

drug treatment

1 1

0 0

0 0

38

Overall, the manual inspection of data for data flaws did not reveal any patterns of errors that

would be a cause of concern: there were no undocumented codes, out of range values and

missing observations were few.

Assessment of the overall data quality: The first part of the critique consisted of evaluating the

database purpose, content, sampling, comparison of the sample to the target population and data

collection methods. The purpose and content of the database were concluded to be a good fit

with the research question for two reasons. First, the purpose of the database was to collect data

in order to help improve the current OP care in the province and the purpose of the research

question was to identify factors which could also be used to improve current OP interventions.

Second, even though the secondary use of already collected data prevented exploration of

additional predictors, most variables of interest were present. The sampling strategy and

comparison of the sample to the target population revealed that the two were similar on all,

except for three variables where slight differences were found. Data collection procedures and

ongoing quality assurance were put in place to monitor data quality across sites and rectify any

issues. The last two parts of the critique showed that the data were of high quality with no

structural inconsistencies and a few missing observations. The quality of the data was evaluated

on a regular basis, as part of quality assurance: if any issues were found, steps were immediately

taken to rectify them. Overall, the provincial database used in this thesis was a unique, valuable

source of information that provided rich, high-quality data that I was able to use to answer the

research question appropriately.

3.4. Chapter summary Data used to address the second objective of this thesis, the OP-fracture link, were derived from

the database created as part of the Fracture Clinic Screening Program, MOHLTC-commissioned

province-wide initiative implemented by Osteoporosis Canada. In order to establish the

suitability of the database to answer the research question, the database purpose, content,

structure and quality were examined and a manual examination of raw samples of data was

performed. The results showed that the database was a fitting source of information for the

research question to be answered. Although some limitations were noted (inability to explore

additional predictors, missing data for those not followed up with, possibility of systematic

39

errors), the strengths and opportunities afforded by this database exceed its limitations. The

database remained the most suitable source of data for my research question.

40

Chapter 4 Factors Associated with Patients Making the Link Between Their

Fragility Fracture and Osteoporosis

4.1 Introduction OP is an asymptomatic disease which often only presents clinically after a fragility fracture

occurs (Brown et al., 2006). In addition to indicating a possibility of compromised bone health,

fragility fractures are strong predictors of future fractures (Klotzbuecher et al., 2000). Many

fragility fracture patients, however, do not perceive themselves to be at an increased risk for

future fractures or that their fracture may be linked to OP (Giangregorio et al., 2008; Meadows et

al., 2005; Meadows, Mrkonjic, Lagendyk, & Petersen, 2004; Meadows & Mrkonjic, 2003;

Gerend, Erchull, Aiken, & Maner, 2006; Fiandt, Pullen, & Walker, 1999). This disconnect is

concerning because fragility fracture patients who do not believe that they are at risk for OP or

future fractures are not likely to perceive OP management as applicable or important to them and

are therefore not likely to engage in it (Giangregorio et al., 2009; Meadows et al., 2005). In

interviews with female fragility fracture patients, Meadows and colleagues (2003, 2004, 2005)

found that patients who did not make the OP-fracture link took little or no action to prevent

another fracture. These patients were not motivated to change their behaviour and reduce risk

because they perceived their fracture as an isolated incident, attributable to external factors

(Meadows et al., 2005; Meadows et al., 2004; Meadows & Mrkonjic, 2003). In our previous

research that used structural equation modeling, we found that perceived need (perception of the

OP-fracture link and perception of bones as “thin”) was the best predictor of initiating first-line

OP treatment (Beaton et al., 2009). Our qualitative research also pointed to the importance of

patients making the OP-fracture link (Beaton et al., 2007): focus group participants described the

link as the pivotal step towards initiating OP management (testing and treatment as appropriate).

Similar to making the OP-fracture link, the concept of perceived susceptibility to OP has been

associated with the acceptance of OP therapy (Cline et al., 2005; Cadarette, Gignac, Jaglal,

Beaton, & Hawker, 2009) and has also been shown to be a predictor of follow up with a

physician for investigation of OP (Bliuc et al., 2006).

41

Previous research examining factors that were associated with patients’ perception of

susceptibility to OP and the OP-fracture link was mostly cross-sectional or qualitative in nature

(Meadows et al., 2005; Meadows et al., 2004; Meadows & Mrkonjic, 2003; Giangregorio et al.,

2008; Giangregorio et al., 2009), making the direction of the association unclear. One such cross-

sectional study involving fracture patients found that OP diagnosis was associated with patients

making the OP-fracture link (Giangregorio et al., 2008), while another study involving

community-dwelling women, found that the use of antiresorptive medication was associated with

high perceived susceptibility to OP (Cline et al., 2005). Other factors found to be associated with

higher perceived susceptibility to OP include younger age (Giangregorio et al., 2008), female

gender (Doheny, Sedlak, Estok, & Zeller, 2007), family history of OP (Gerend et al., 2006) and

more OP knowledge (Doheny et al., 2007). Apart from being cross-sectional, previous studies

included patients who were already diagnosed and/or treated for OP (Cline et al., 2005;

Giangregorio et al., 2008), thus making it more difficult to determine which came first: OP

diagnosis/treatment or the perception of the link.

Although previous research has identified some factors associated with fragility fracture patients’

perception of the OP-fracture link and their susceptibility to OP, its cross-sectional nature makes

it impossible to predict which undiagnosed, untreated patients are more likely to make this link

after sustaining a fragility fracture. The objective of this study is to fill this gap in knowledge by

identifying baseline predictors that characterize undiagnosed and untreated patients who make

the OP-link at follow up. Identification of factors predictive of (rather than associated with)

making the OP-fracture link can help identify specific groups of patients at the time of initial OP

screening, when intervention can be delivered. Patients identified as less likely to make the OP-

fracture link might require a different, more intense intervention to make this link. The ability to

predict which patients will (and will not) associate their fracture with OP would therefore: (a)

help identify the patients who need a modified intervention and (b) help create an appropriate

intervention tailoring it to the patients less likely to make the link.

Since only two studies examined the OP-fracture link (Giangregorio et al., 2008; Meadows et

al., 2005), most predictors were based on research examining a related concept, patients’

perceived susceptibility to OP and future fractures (Doheny et al., 2007; Gerend et al., 2006).

Other predictors were drawn from our previous research on the pathway to OP care (Beaton et

al., 2007). Based on the previous research, we hypothesized that patients more likely to make the

42

link at follow up would be: younger, female, those who sustained a more complex fracture (hip,

shoulder or multiple fractures), those with a previous fragility fracture after the age of 40, as well

as maternal fracture (also after the age of 40). We also hypothesized that the following

modifiable factors would be associated with making the OP-fracture link: perception of bones as

“thin”, having more knowledge of OP and perceiving more benefits and fewer barriers to OP

treatment.

4.2 Methods Sample: The sample for this study was drawn from the database that was created as part of a

provincial screening program for OP. At the time data were collected, there were 19 screening

coordinators placed in 36 designated fracture clinics across the province in order to identify,

assess, refer and educate fragility fracture5 patients over the age of 50 (or younger if referred by

a surgeon). Screening coordinators collected self-reported patient data at baseline and follow up.

The following data were collected at baseline: OP risk factors, OP diagnosis, bone mineral

density (BMD) testing and OP treatment history, demographics, documentation of

intervention/patient consent and patient’s OP knowledge and beliefs. Patients who were not

diagnosed or treated for OP at baseline were contacted at three or, if necessary six months6 and

asked about post-fracture follow up with their family physician and new testing, diagnosis and

treatment information in addition to the same OP knowledge/beliefs questions that were asked at

baseline.

The study sample included consenting patients who were not diagnosed or treated for OP, whose

baseline data were completed by September 30, 2009 and who were followed up by December

31, 2009. For the purpose of answering the research question, the sample was further restricted to

those patients who, at baseline, did not think that their fracture was related to OP. This study is a

part of a thesis project which received ethics approval from St. Michael’s Hospital and the

University of Toronto research ethics boards (Appendix IV). The research database from which

5 Fragility fracture was defined as a fracture that resulted from a fall from standing height or less 6 Patients who have not followed up with their GP by three months post-fracture were contacted again at six months

43

the study sample was drawn was approved by the research ethics boards of each participating

site.

Measures: The study’s outcome variable was the OP-fracture link as measured by a “yes”

response to the following question “Do you think your broken bone could have been caused by

having OP (thin or brittle bones)?” Contrasting responses (“no” and “do not know”) were

collapsed into one category “did not make the link”. This was done because the purpose of the

research question was to distinguish only between those aware of the OP-fracture link (“yes”)

and those unaware of this link (“not yes”). Also, the sample only included those patients who

answered the OP-fracture link question at both baseline and follow up.

The variables chosen as potential predictors are shown in figure 4.1 and listed below:

1. Patient’s age at the time of screening: <65 vs. ≥65. Patients’ age was dichotomized into

these categories because patients over the age of 65 are at a higher risk for OP (Brown &

Josse, 2002; Brown et al., 2006) and because the division into <65 vs. ≥65 provided us

with two groups with similar number of observations: n=415 and n=438 (65 was the

median).

2. Gender: male vs. female. Gender was chosen as a predictor because previous research has

not addressed the question of whether there was a gender difference in who is more likely

to make the link.

3. Fracture site: shoulder, hip and multiple vs. single fractures. Our previous qualitative

research showed that patients who made the OP-fracture link also reported having a more

complicated fracture, such as a hip or a shoulder fracture (Beaton et al., 2007). We

therefore divided all patients into two groups: those who had a shoulder, hip or a multiple

fracture and those who sustained a single fracture.

4. Previous fragility fracture after the age of 40: yes vs. no (as measured by a yes/no answer

to the question “Has the patient ever broken any other bones since the age of 40 from a

simple trip and fall?”). Previous fracture was associated with awareness of bone health in

previous research (Giangregorio et al., 2008).

44

5. Maternal history of fractures: yes vs. no (as measured by a yes/no answer to the question

“Did this patient’s biological mother have any fractures after the age of 40?”). A 2006

study found that women who rated their risk of developing OP as high attributed this risk

primarily to family history (Gerend et al., 2006).

6. Perception of bone quality: thin vs. do not know vs. normal. Perception of one’s bone

health was measured by patient’s agreement with the statement “My bones may be

thin/my bones are definitely thin”. This variable was chosen because previous research

pointed to the lack of awareness of poor bone health in most fragility fracture patients

(Meadows et al., 2005; Meadows et al., 2004; Meadows & Mrkonjic, 2003).

7. Knowledge of OP. OP knowledge was associated with perceived susceptibility to OP in

previous research (Doheny et al., 2007). We measured the OP knowledge using four

items from the “OP and You” knowledge questionnaire (Brenneman et al., 2002): “There

is no way to prevent OP”, “Bones cannot be rebuilt once they are thin from OP”, “If a

person has OP, something as simple as lifting a bag of groceries can break a bone” and

Outcome: “My fracture

may have been caused by OP”

Initiation of OP

management

OP therapy benefits outweigh

barriers

Quality of bones -

thin

Previous fracture

Younger age

Fracture site: more painful

fracture

Maternal fracture

Female gender

Figure 4.1 Hypothesized predictors of making the osteoporosis-fracture link by follow up

More OP knowledge

45

8. Perceptions of benefits and barriers to OP pharmacotherapy. These variables were

included as predictors because our previous qualitative research found that patients’

health beliefs played a role in making the OP-fracture link. Perceived benefits of OP

pharmacotherapy were measured by a five-item OP benefits scale while the perception of

barriers to OP pharmacotherapy was assessed by two individual items. Both the

perceived OP benefits scale and the items measuring OP barriers were derived from

previous work by Cadarette (Cadarette et al., 2004) who based her scale and items on the

Osteoporosis Health Belief Scale (Kim et al., 1991). The number of perceived benefits

was measured on a scale 1-5 while barriers were assessed individually (yes/no to each of

the two questions).

Analysis: The sample was first described on a univariate level. Cronbach’s alpha coefficients of

the OP knowledge, OP pharmacotherapy benefits and OP pharmacotherapy barriers questions

were examined. The sample was compared to the population from which it was drawn to

examine any potential systematic sample biases. Distribution of the outcome variable (OP-

fracture link at follow up) was assessed and dichotomized into two responses: making the OP-

fracture link (“yes”) and remaining unaware of this link (“not yes”). Bivariate (unadjusted)

analysis was used to examine the relationship between each predictor and the outcome (chi

square tests for categorical and 2-tailed t tests for continuous variables). Multivariable logistic

regression was then used to determine which baseline factors were associated with the OP-

fracture link at follow up while controlling for other predictors. As per Hosmer and Lemeshow

(Hosmer & Lemeshow, 2000) only predictors with a p-value of <0.25 (based on unadjusted

analysis) were used in model building. Multicolinearity was assessed by examining the tolerance

and variance inflation factors of each variable to determine whether or not any of the predictor

variables were correlated. Tolerance levels below 0.4 and variance inflation values greater than

46

2.57 were to be taken as an indication that predictors were highly correlated8 (Allison, 1999).

Manual backwards elimination was then used to reduce the number of predictors in the final

model (Harrell, 2001): it was decided a-priori that a non-significant predictor would be kept in

the model if it changed the estimate of other predictors by 20% when taken out. Manual

backwards elimination allowed us to closely monitor the impact of the removal of each variable

on the model as a whole and on other predictors. The maximum number of predictors in the final

model was limited to one per ten events in the smallest category. Goodness of fit was assessed by

Hosmer and Lemeshow’s Goodness-of-Fit Test and accuracy of the model as a predictor of

outcomes by concordance statistic (Hosmer & Lemeshow, 2000). All analyses were performed

using SAS Version 9 (Statistical Analysis System, Cary, N. Carolina). The criterion for

interpreting statistical significance at the final stage was set at alpha ≤ 0.05.

4.3 Results Psychometric analysis: Cronbach’s alpha coefficient for the two OP knowledge domains that

the OP knowledge questions tapped into confirmed previous findings (Cadarette et al., 2007) of

low internal consistency of the two domains (Cronbach’s alpha coefficients of 0.43 and 0.24).

We therefore analysed the four OP knowledge questions as separate items, dichotomized into

“correct” and “incorrect”. Cronbach’s alpha coefficient for questions tapping into the patient

perception of OP benefits and barriers showed that the internal consistency for the OP benefits

scale was acceptable for group level analyses such as ours (Cronbach’s alpha coefficient of 0.76)

but was low for the OP barriers domain (Cronbach’s alpha coefficient of 0.19). Based on this

analysis and previous recommendations (Cadarette et al., 2004), the perception of OP

pharmacotherapy benefits was analysed as a scale, while the two barriers items were analysed

separately as a dichotomy of correct and incorrect answers.

7 High inflation value indicates that the variance of a regression coefficient is increased because of collinearity; low tolerance value also indicates that the variable under consideration is highly correlated with another variable already in the model. 8 In case of multicollinearity, a less important and less modifiable variable would be taken out of the model.

47

Analyses examining OP-fracture link: Of 916 patients who had data available on the OP-

fracture link question (at both baseline and follow up), 93% (n=853) did not make this link at

baseline. The distribution at follow up was similar: of those patients who did not make the link at

baseline, 91.2% (n=778) still did not make the link or remained unsure about it (Figure 4.2).

Figure 4.2 Distribution at follow up of those who did not make the OP-fx link at baseline

At baseline: 853 (93% of 916) patients did not make a link between OP

and their fracture (or were unsure about it)

At follow up:

75 (8.8%) patients made the link by follow up

At follow up:

778 (91.2%) patients still not making the link

In the sample used for the analysis (patients who did not make the link at baseline, n=853), the

majority of patients were female (80%) with the mean age of 67 (SD =10.5, range 50-94).

Twenty-five percent of the patients sustained a hip, shoulder or a multiple fracture while 28%

reported a history of maternal fractures (after the age of 40). The baseline fracture was not the

first fragility fracture after the age of 40 for 26% of the patients. This sample was compared to

the population it was drawn from on variables of interest: potential confounders, predictors and

the outcome. The results are shown in Table 3.1

As compared to the target population, the sample consisted of fewer patients who were employed

(30.1% vs. 35.4%), a greater number of patients who perceived more benefits of OP drug

treatment (4.2 of 5 and 4 of 5), more oral steroid users (6.3% vs. 4.3%) and more patients who

correctly answered the knowledge question stating “If a person has OP, something as simple as

lifting a bag of groceries can break a bone” (84% vs. 78%). Although statistically significant, the

differences in perceived treatment benefits or steroid use were not substantial: both groups

perceived four of five benefits on average and there was a 2% of difference in steroid use

between the two groups. The sample and the population did not differ on other variables of

interest. Specifically, there were no statistically important differences in age, gender, marital

status, living situation, type of fracture, existence of previous or maternal history of fractures,

48

previous falls, feeling unsteady, perception of the bone quality, the remaining knowledge items

(three of four) and perceived barriers to OP drug treatment.

Table 3.1 Description and comparison of the sample n=853 to the target population n=2237 Variables n=853 n=2237 p value

Mean age (SD) range 66.6 (10.5) 50-94

66.6 (10.8) 50-95 0.97

Gender – Female

79.7%

76.8%

0.08

Marital status: Married/common law

59.7%

62.5%

0.34

Living situation: Live alone

28.7%

26.8%

0.31

Work full or part time

30.1%

35.4%

0.0054

Hip or shoulder fracture (may include others)

25.6%

24.5%

0.56

Previous fracture >age 40

25.2%

25%

0.94

Maternal fracture >age 40

28%

25%

0.09

Previous fall in the last year

26.6%

27.2%

0.73

Feeling unsteady

27%

30.2%

0.08

Current or past steroid use

6.33%

4.31%

0.02

Perceived benefits of OP drug treatment (1-5)

4.2

4

0.0032

% Who perceived bones as thin

10.6%

10%

0.11

Knowledge item: There is a way to prevent OP (reversed)

64.2%

67.7%

0.07

Knowledge item: Bones can be rebuilt once they are thin from OP (reversed)

57.1%

55.5%

0.41

Knowledge item: If a person has OP, lifting a bag of groceries can break a bone

84.3%

78.7%

0.0006

Knowledge item: Health problems caused by OP can be life-threatening

69.2%

71.1%

0.33

Perceived barriers item: Taking too many medications

22.2%

20.8%

0.38

Perceived barriers item: Stomach problem limits ability to take drug treatment

17.5%

20.3%

0.1

49

Logistic regression was used to examine which of the hypothesized baseline factors were

associated with making the OP-fracture link at follow up. Of the 13 hypothesized factors

(including four knowledge items and two barriers), nine predictors had a p < 0.25 based on the

unadjusted analysis. These factors were entered into the adjusted (multivariable) analysis:

manual backwards elimination was used to remove the least significant variable sequentially to

narrow down the number of predictors, while examining the change in estimates in the remaining

variables. The final model contained six predictors (maximum number of predictors for the

number of events). P values associated with Type I error of 0.05 (p<0.05) were considered

statistically significant. The results of the unadjusted and adjusted analyses are presented in

Table 4.1.

Table 4.1 Regression of baseline factors most likely to be associated with a transition to making the link

Baseline characteristics: Unadjusted OR (95% CI) p value Adjusted OR

(95% CI) p value

Age < 65 vs. ≥ 65 1.5 (0.91-2.4) 0.12 dropped* -

Gender: female vs. male 1.96 (0.95-4.0) 0.06 1.8(0.8-4.1) 0.18

Previous fracture > age 40 yes vs. no 2.6 (1.6-4.2) <.0001 2.2 (1.3-3.9) 0.004

Maternal fracture >age 40 yes vs. no 1.4 (0.9-2.3) 0.18 dropped* -

OP knowledge item: Bones can be rebuilt once they are thin from OP (correct vs. incorrect)

1.43 (0.9-2.4)

0.16 1.5 (0.8-2.9) 0.2

Perceived OP pharmacotherapy benefits (yes vs. no)

1.3 (1.0-1.8) 0.03 1.2 (0.9-1.6) 0.13

Perceived barrier to OP pharmacotherapy Patient cannot take more drugs due to stomach problems (agree vs. disagree)

0.6 (0.3-1.3) 0.2 1.6 (0.8-3.5) 0.2

Perception of bone quality: • Do not know what the bone quality is • Bones thin or thinning • Normal or normal for age

3.4 (1.9-6.1) 9.8 (5.3-18.0)

ref

<.0001 <.0001

-

2.8(1.5-5.4) 7.1(3.7-13.8)

ref

0.0016 <.0001

-

*Dropped = dropped during the adjusted analysis

50

In the unadjusted analysis, a previous fragility fracture, high perceived benefits of OP

pharmacotherapy and perception of bones as thin or ambiguity about the bone quality were all

associated individually with making the OP-link at follow up. Perception of benefits was not

significant in the adjusted analysis which showed that, controlling for other factors, the odds of

making the link at follow up were: 2.2 times higher for those who sustained a fracture prior to

baseline fracture, 2.8 times higher for those who were unsure about the quality of their bones at

baseline (vs. those who thought bones were normal) and 7.1 times higher for those who believed

at baseline that their bones were thin (vs. those who thought the bones were normal). This model

showed good fit with the data: the Hosmer and Lemeshow Goodness-of-Fit Test resulted in a p

value of 0.63, while the model accuracy was confirmed by the concordance statistic of 0.76.

4.4 Discussion The results of this study indicate that most fragility fracture patients do not associate their

fracture with OP, even after a post-fracture intervention. This study identified the baseline

characteristics predictive of making the OP-fracture link at follow up: those who became aware

of the OP-fracture link by follow up were more likely to have had a previous fracture, perceive

their bones as thin or be unsure about their bone quality.

The findings from our study confirm the results of previous research on the OP-fracture link

which showed that most fragility fracture patients do not link their fracture with underlying OP

(Giangregorio et al., 2008). A novel finding from our study is that, for most patients, this

perception does not change over time: 91% of patients who did not make the link at baseline do

not make this link at follow up either. This is concerning because a majority of fragility fracture

patients in the orthopaedic environment have a low BMD (Astrand et al., 2008) and are in need

of OP management (Lindsay et al., 2001; Astrand et al., 2006a; Center et al., 2007) but they are

not likely to initiate if they do not find OP management to be applicable to them (Giangregorio et

al., 2008; Meadows et al., 2005).

Only two of the 13 hypothesized baseline factors were found to be predictive of the OP-fracture

link at follow up. Patient perception of their bones as “thin” and a previous fragility fracture

were mentioned in previous research as being correlated with perceived susceptibility to OP and

51

future fractures. Meadows (2005) found that the only group of fragility fracture patients that took

action towards managing their OP was the group that perceived their bone health as

compromised and themselves at risk for future fractures. In addition to the perception of thin

bones, previous research found that having more than one fragility fracture was related to the

perception of susceptibility to OP (Giangregorio et al., 2009): this study found that having one

fracture was not necessarily associated with the awareness of bone health but more than one was.

Having more than one fragility fracture was found to be associated with the OP-fracture link in

our previous focus group research; however, not everyone associated their fractures with thin

bones: one patient thought she was just “clumsy” and was repeatedly fracturing due to external

circumstances, not compromised bone health (Beaton et al., 2007).

Surprisingly, factors such as gender, age, OP knowledge and perceived benefits of OP

pharmacotherapy were not found to be predictive of the OP-fracture link at follow up. Women

were no more likely than men to make the OP-fracture link, a finding that was surprising in light

of previous cross-sectional research on perceived susceptibility to OP that identified women as

more likely to make the link (Doheny et al., 2007). This finding also does not reflect the fact that

women are at a higher risk of fragility fractures and OP (Kanis & Johnell, 2005; Tenenhouse et

al., 2000). In previous research examining the association between age and future fracture risk

(Giangregorio et al., 2008), decreasing age was found to be related to the perception of future

fracture risk which is particularly concerning because increasing age is actually an independent

predictor of subsequent fracture (Kanis et al., 2004). Our study demonstrated that age was not a

significant predictor of making the OP-fracture link between baseline and follow up, however,

we did not assess whether age was a predictor of perception of future fracture risk. The findings

of our study indicate that gender and age may only be correlated with concurrent perception of

OP-fracture link and/or susceptibility to OP but are not predictive of becoming aware of this link.

OP knowledge at baseline (knowledge of OP treatments and consequences of OP) was also not

related to making the OP-fracture link at follow up. This finding was unexpected given that

previous research found that those who knew more about OP also thought they were more

susceptible to it (Doheny et al., 2007). Previous research using structural equation modelling

(Beaton et al., 2009) also showed that baseline knowledge of OP was indirectly associated with

making the OP-fracture link at follow up. The analysis showed that previous OP knowledge,

along with the perception of OP treatment benefits, was strongly associated with OP knowledge

52

(and perception of treatment benefits) at follow up which, in turn, was associated with a

combined variable called “perceived need” consisting of the perception of thin bones and the

OP-fracture link at follow up (our OP-fracture link analysis, however, kept the two variables

separate). It is also possible that baseline knowledge alone may not predictive of the OP-fracture

link but that, in conjunction with OP benefits, it may indirectly impact making of the link at

follow up.

The current study is unique because it is a first cohort study that examined predictors of the OP-

fracture link in fragility fracture patients who were not diagnosed or treated for OP and who also

were unaware of the OP-fracture link at baseline. Previous research, due to its cross-sectional

nature, was only able to identify factors that were associated with the link and not predictive of

it. By determining which factors were predictive of OP-fracture link, our study also established

which factors were not predictive of the link, an equally important finding. An important strength

of the study rests with the provincial data that the study was based on: having data from multiple

sites across the province ensured that our sample was representative of the fragility fracture

population to which the results were to be generalized. Lastly, limiting our sample to

undiagnosed and untreated patients allowed us to eliminate the possibility of patients making the

OP-fracture link because of their diagnosis/treatment recommendations. Associating OP and

fractures is important because OP is an asymptomatic disease: patients who are unaware of their

risk factors are unlikely to take any action to prevent bone disease or to prevent already

compromised bone health from further deteriorating.

The main limitation of our study is the reliance on data which were not specifically collected for

the purpose of answering the research question on the OP-fracture link. We were therefore

unable to include and examine more potential predictors of interest or to base data collection on a

specific model of determinants of health behaviour change. Another limitation associated with

the database is a possibility of sampling bias as our sample and the target population were

different on three variables.

The findings of this study could be incorporated in post-fracture OP interventions to better

communicate to patients their risk of future fractures and OP. Patients who, at baseline, did not

have concerns about their bone quality and those who did not have a previous fracture are at a

particular risk for not making the OP-fracture connection. Future interventions could target these

53

patients (who did not make the link at baseline) emphasizing the link between a previous fracture

and OP and delivering a clear message about the quality of patient’s bones as “thin”. The same

message should be reinforced in patients who have experienced a previous fracture and who

already have a perception of their bones as “thin” in order to link the two with the need for

initiating OP management. Future post-fracture interventions should therefore support the

awareness in the patient group likely to make the OP-fracture link and use a different approach in

those not fitting this profile. Future research should explore the impact of such interventions on

acceptance and adherence to recommended therapies to reduce fracture risk.

54

Chapter 5 Discussion and Conclusion

5.1 Discussion The objectives of this thesis were twofold: to explore the extent to which fragility fracture

patients’ perception of their bone health has been considered in existing post-fracture OP

interventions and to examine fragility fracture patients’ awareness of their bone health in relation

to the fracture they sustained. The ultimate goal of the thesis was to inform the creation of more

effective, patient-centered OP interventions that are based on patients’ perception of their bone

health.

This thesis focused on the patient and patient-related barriers to more effective OP management

because it is ultimately the patient who needs to implement and adhere to the recommendations

and prescribed therapy. For OP management to take place, patients need to make changes to their

existing behaviour (e.g., increase their Calcium intake) or adopt a new behaviour altogether (e.g.,

initiate bone sparing medication). Whether such behaviour change will occur largely depends on

patients’ perception of the state of their health and perceived health risks: it is known that

patients’ perception of their health has a large influence on health-related behaviours including

treatment decisions (Gold et al., 2006; Deeks, Zoungas, & Teede, 2008). Patients’ perception of

their health and the process of behaviour change are concepts central to individual-level theories

of health behaviour change. This focus on patients’ perception distinguishes theories of health

behaviour change from educational interventions that provide generic information to all patients

without taking into consideration patients’ perception of the issue at hand. Application of

theories of health behaviour change takes into account the differences in patients’ perceptions

and strives to adapt the health information to each patient. These theories have been applied to a

variety of contexts that involve stopping, initiating or maintaining specific behaviours such as

smoking cessation, weight reduction, exercise acquisition (Glanz et al., 2002), chronic disease

management such as rheumatoid arthritis (Mulligan & Newman, 2003), asthma (Becker et al.,

1978)) and OP prevention (Gold & Silverman, 2004). Given the previous successful application

of theories of health behaviour change to a variety of behaviours and the importance of patients’

55

perception of their health in treatment initiation, the following question arises: “Have theories of

health behaviour change been used to help fragility fracture patients change their behaviour and

initiate OP management (testing and treatment as appropriate)?”

The first objective of the thesis was to examine whether behaviour change theories (or any of the

concepts found in these theories) have been integrated into post-fracture OP interventions. A

review of patient-targeted, primary studies that aimed at improving post-fracture OP care was

undertaken. The results showed that theories of health behaviour change have not been used in

post-fracture interventions and that patients’ perception of their bone health has not been

incorporated in current post-fracture OP management. Only three of the 35 reviewed studies

even mentioned constructs suggestive of behaviour change theories. One study (Vaile et al.,

2007) was consistent with the “teachable moment” heuristic, describing the fracture as an

opportunity to successfully intervene and motivate subsequent behaviour change. The second

study (Davis et al., 2007), targeting patients’ awareness of OP and encouraging them to take on a

more active role in OP management, was suggestive of another health behaviour change model,

Social Cognitive Theory (Bandura, 2004). The third study (Astrand et al., 2006; Astrand et al.,

2008), built on an assumption that patients would make rational decisions based on the

information provided, was a fit with the Health Belief Model (Rosenstock et al., 1988), another

major theory of health behaviour change. Each of these theories (and other major theories of

health behaviour change) could be formally applied to the context of OP and incorporated in

post-fracture OP interventions. The interventions based on theories of behaviour change have a

potential to be effective because of the emphasis that these theories place on patients’

perceptions of their health, the importance that patients’ perceptions have in the actual behaviour

change and the fact that the theories and relationships between the constructs in them have been

established based on existing knowledge (Glanz et al., 2002).

Focusing on an individual patient’s behaviour is particularly important in OP management

because the current healthcare environment is filled with barriers to more effective management

of this condition (Simonelli et al., 2002; Kaufman et al., 2003). In such an environment, there is

a big need for the patient to be an agent of change. If the patients are to take on this role, they

first need to modify their own behaviour. In order to help patients change their own behaviour

successfully, we need to explore patient-related factors and barriers, particularly the ones

associated with patients’ perceptions that impact their behaviour.

56

The second objective of the thesis further explored patients’ perception of their bone health,

specifically focusing on patients’ perception of the link between their fracture and an underlying

bone disease, OP. This link is significant because a fragility fracture is usually the first (and often

the only) sign of compromised bone health and is an opportunity to make the issue of

compromised bone health salient for patients. This is especially important given the otherwise

asymptomatic nature of OP: if patients have no symptoms indicating that their health is

compromised, why would they do something about it?

Findings derived as part of the second thesis objective showed that most patients do not make the

link between their bone health and fragility fracture and that this perception does not change over

time in most cases: 91% of patients who did not make this link at baseline, do not make this link

at follow up either, even after a baseline intervention. This is concerning because the majority of

fragility fracture patients (in the orthopaedic environment) require OP management (Astrand et

al., 2006) and yet, most of these patients may not perceive a rationale for OP management

because they do not perceive the link between their fracture and OP.

The second objective also identified factors that characterized patients likely to become aware of

the OP-fracture link at follow up after having been unaware of this link at the time of screening.

Factors that we theorized could be predictive of the link were chosen based on previous research.

Only two of 13 theorized predictors were found to be predictive of making the OP-fracture link

at follow up: perception of bones as “thin” and having a previous fragility fracture (at baseline).

Equally important, the results also showed that age, gender, a history of maternal fractures and

perception of OP treatment benefits and barriers were not predictive of making the OP-fracture

link at follow up. Knowing which factors were not predictive is important because some of them

(age, gender etc) have been found to be related to perception of susceptibility in previous

research. Since the results were obtained by utilizing logistic regression, predictors of making

the OP-fracture link at follow up revealed the predictors of not making the link at follow up. This

allowed for identification of patients who are at a higher risk for not making the link and who

may require a different approach to intervention than those who do make the OP-fracture link.

5.2 Key findings The results of the first thesis objective indicate that behaviour change theories, which have the

potential to improve current OP management, have not been incorporated into existing post

57

fracture OP interventions. The results of the second thesis objective demonstrate that most

patients do not perceive the OP-fracture link, that this perception does not change at follow up in

most cases and that patients without a prior history of fractures and those who do not perceive

their bone as “thin” are particularly at risk for not making the OP-fracture link at follow up, even

after an educational intervention at baseline. Taken together, the results point to the need to (a)

try a different approach that accounts for the patient perception of the OP-fracture link, such as

the approach suggested by theories of health behaviour change and (b) to create OP interventions

for a specific group of patients (those who did not make the OP-fracture link at baseline) who are

not likely to make this link by follow up either,. This intervention should emphasize that

patients’ bones are “thin” and that even one previous fragility fracture is indicative of OP.

5.3 Implications The review of post-fracture OP interventions revealed that there is an untapped source of

knowledge that can be used to improve the existing OP interventions, namely, health behaviour

change research. Application of models of health behaviour change would help explain the

process of patients’ initiation of OP management and as such, could contribute to improving

current OP interventions.

Five major individual-level theories of health behaviour change were reviewed in Chapter 2: the

Health Belief Model (Rosenstock et al., 1988); Social Cognitive Theory (Bandura, 2004);

Teachable Moment (McBride et al., 2003); Theory of Planned behaviour (Fishbein M & Ajzen I,

1975) and the Transtheoretical Model (Prochaska & Velicer, 1997) (Table 1). Even though each

one of these theories could be applied to the context of post-fracture OP management, the results

of the second objective point to some of the theories being more applicable to the OP context

than others. The finding that most patients do not make the link between their fracture and OP

and that this perception, in most cases, remains constant over time indicates that most patients

may not perceive a reason to initiate OP management because they do not see themselves as

susceptible to OP as a health threat. Theories that do not incorporate the concept of patients’

perceived susceptibility may not be as applicable to OP because these theories presume that the

threat is already perceived. Social Cognitive Theory and the Theory of Planned Behaviour, for

example, include concepts of self-efficacy, perception of benefits and barriers to behaviour

58

change, proximal goals and the impact of social norms, but they do not account for patients being

unaware of their health risk to begin with and the need to change their behaviour. The

Transtheoretical Model also does not include the concept of perceived susceptibility to a

condition but it does describe a stage (precontemplation) where patients are not necessarily

aware of their susceptibility to a condition and/or where patients do not consider their existing

behaviour as problematic. The lack of perceived susceptibility that exists in the context of OP

could be addressed in the precontemplation stage of the Transtheoretical Model if the

consciousness-rising process, specific to this stage, was adapted to include more emphasis on

making the OP-link and increasing the awareness of susceptibility.

Perception of susceptibility to a condition/health threat is one of the main concepts in the Health

Belief Model and the Teachable Moment. These theories are not based on an assumption that

patients necessarily perceive a health threat or their existing behaviour as needing modification.

For this reason, these theories may ultimately be more applicable to the concept of OP. Most

fragility fracture patients do not perceive themselves as susceptible to OP based on their fracture

and, consequently, do not perceive a need for behaviour change. The Health Belief Model and

the Teachable Moment could therefore be used to explain (in part, at least) the low rate of OP

management: patients are not likely to take any action if they do not even perceive themselves as

susceptible to a condition.

The Teachable Moment “cueing event” heuristic can also be used to explain why so few patients

initiate OP management. According to this heuristic, whether an event is significant enough to

become a “teachable moment” depends on (a) whether it increases the perception of risk, (b)

whether it prompts an affective response, and (c) whether it leads to a re-definition of a health-

related self-concept (McBride et al., 2008; McBride et al., 2003). In the context of OP, a fragility

fracture has the potential to become a “teachable moment” but, for most patients, this does not

happen because (a) patients do not associate their fragility fracture with OP as a health threat, (b)

the fracture does not prompt an affective response that would motivate patients to initiate OP

management because of the lack of the OP-fracture link (the fracture may prompt the patient to

initiate fracture management instead), and (c) no re-definition of self concept/social role

happens, again, because of this lack of the OP-fracture link.

59

Post-fracture OP interventions could be based on the Health Belief Model and the Teachable

Moment frameworks. First, fragility fracture patients who do not perceive themselves as

susceptible to OP would need to be identified in order to help these patients make the link

between their fracture and the possibility that they may have OP. In order to help patients make

the link, predisposing factors (such as a previous history of fractures, patients’ beliefs about OP

etc.) would be assessed and taken into consideration as these may impact patients’ cognitive

reaction to the cueing event (fracture). A clear message about the quality of patients’ bones as

“thin” would be delivered to these patients emphasizing that even one fragility fracture is

indicative of an underlying bone disease. Making the link between OP and the fracture is crucial

for the application of the Teachable Moment if the fracture is to be used as a cueing event. By

making the fracture a relevant “cueing event”, linking it to the possibility of OP, patients’

perception of risk of OP is likely to increase. Perception of the fracture as a threatening event

that increases the risk of OP may, in turn, impact another component of the Teachable Moment

heuristic, affective response. An affective response, such as worry, contributes to creating a

“teachable moment” and may motivate patients to take action to reduce the risk and the resulting

affective response (McBride et al., 2008). The last component of the Teachable Moment

heuristic, health-related self-concept, could also be impacted by establishing the OP-fracture link

and thus making the fracture relevant. A possibility of OP may be perceived as a threat to current

or future self-concept and may prompt behaviour change. The self-concept, however, is also

impacted by external factors such as the influence of peers and social norms. Compared with

many other chronic conditions (e.g., diabetes, heart disease) where patients get a fairly consistent

message from most sources such as health care practitioners or peers (Persell et al., 2010;

Boothroyd & Fisher, 2010; Funnell, 2010) the message that fragility fracture patients get is not

as consistent and, in many cases, is completely absent as many health professionals do not see

addressing OP as their responsibility, or they simply may lack time to address it (Kaufman et al.,

2003; Levinson & Clay, 2009; Boden et al., 2005). The perception of OP as a normal part of

aging that affects elderly women, rather than as a treatable chronic disease that can happen to

anyone, may reinforce the message that the patient is not at any risk for OP and therefore, does

not have a reason to take action. In such an environment, a clear, consistent message about the

importance of bone health and the link between OP and fractures could re-affirm the significance

of the fracture and increase the chances that fracture will become a “teachable moment” that will

motivate patients’ behaviour change.

60

Establishing the OP-fracture link, and making the fracture as significant a cueing event as

possible, are initial (and necessary) steps in the application of the Health Belief Model and the

Teachable Moment to post-fracture OP management. Once the OP-fracture link is made and the

fracture is perceived as a “teachable moment”, other components of these theories can be

incorporated into the intervention. For example, the concepts of self-efficacy, perceived benefits

and perceived barriers, which previous research has shown to be most influential in predicting

behaviour change (Janz & Becker, 1984; Wallace, 2002) should be included in OP interventions.

A fragility fracture presents a unique opportunity for patients to become aware of their

susceptibility to OP because the fracture is the first and (usually) the only sign of compromised

bone health. Sadly, as the results of the second objective of this thesis confirm, most fragility

fracture patients do not perceive the link between their fracture and OP. Theories of health

behaviour change could be very helpful in this context, particularly the ones that would account

for patients’ perception of their susceptibility to OP and patients’ perception of their fracture as a

“cueing event” within a “teachable moment”.

5.4 Strengths There are several strengths of the studies completed as part of this thesis. The literature review

that was performed as part of the first objective was the first review of theoretical frameworks in

the field of post-fracture OP interventions. It was built on the final stage of a systematic review

of non-surgical interventions in the orthopaedic environment and based on 333 articles. The

findings suggest a new, promising direction for future research into improving post-fracture OP

management. In particular, the review suggests that theories of health behaviour change, which

have not been used in post-fracture OP interventions so far, could be applied to this context.

The second part of the thesis, examination of predictors of the OP-fracture link, was based on the

first longitudinal study that examined the change in patients’ perception of the OP-fracture link

over time. This is also the first study (to our knowledge) that excluded patients who were

previously diagnosed and/or treated for OP and who did make the OP-fracture link at baseline.

This study was done with data drawn from a unique, first of its kind source of data, the Fracture

Clinic Screening Program database, which was created as part of Ontario’s first OP strategy. The

61

database included data obtained from multiple locations across the province and collected in the

fracture clinic environment, an environment where the concentration of fragility fracture patients

is the highest. In addition, this database was created to collect data for both program evaluation

and research and it included concepts that would not normally be present in large administrative

databases such as concepts of patients’ OP knowledge and beliefs about bone quality, OP

treatment and barriers to this treatment. The results of the thesis identified factors predictive and

not predictive of making the OP-fracture link, identifying the patient group requiring a different

OP intervention. The findings also contributed to the literature by establishing which factors are

not predictive of making the OP-fracture link even though they may intuitively seem like good

predictors (i.e., patient’s gender, age, history of maternal fractures etc).

5.5 Limitations The main limitation of the literature review is that it was based on the systematic review that

utilized only three major databases (CINAHL, EMBASE, OVID Medline) and therefore did not

include unpublished, non-peer reviewed work/grey literature and conference abstracts. Even

though this approach may have resulted in omission of some potentially relevant theoretical

frameworks, it was the most suitable approach for addressing our research question of whether

the current OP interventions in the orthopaedic setting were informed by theories of health

behaviour change. We specifically focused on peer-reviewed, published studies with primary

data that describe post-fracture interventions in the orthopaedic setting because they include the

setting and the population we are interested in. Including studies done in other settings and with

different populations could have uncovered more theoretically-based interventions, however, we

were interested in studies that guide evidence-based post-fracture practice in the orthopaedic

environment. Furthermore, we did not include studies published in languages other than English.

Another limitation of the literature review is that it only focused on five major individual-level

theories of health behaviour change (the Health Belief Model, the Social Cognitive Theory, the

Teachable Moment, the Theory of Planned Behaviour, the Transtheoretical Model). The purpose

of the review, however, was not to be exhaustive in accounting for use of every theory but to

examine the extent to which patient perception was integrated into post-fracture OP

interventions. Since all individual-level theories of health behaviour change focus on patient

62

perception, we decided to examine interventions for explicit or implicit use of major theories.

We also examined the interventions for the mention or description of any of the concepts (from

these theoretical frameworks) because of the possibility that the word limit set by a journal or the

journal’s aim and scope prevented the author from providing a full description of a theoretical

intervention on which the intervention was based. In our review, we tried to be as sensitive as

possible to any descriptions of concepts that could be related to individual theories of health

behaviour change.

A limitation of the second objective, which involved the analysis of the OP-fracture link, is that

this analysis was based on already collected data and that some concepts of interest (such as the

concept of patients’ health literacy) could not be explored for this reason. This omission,

however, created a topic for future research which may examine the question of whether

patients’ health literacy is predictive of making the OP-fracture link. Other factors not included

in the database are variables tapping into cultural and social issues, co-morbidities and the level

of education. These potential predictors were therefore not available in this analysis.

Another limitation of the OP-fracture link analysis is the possibility of a sampling bias. Since the

sample (n=853) was drawn from an eligible population of n=3090, the two populations were

compared to determine whether there were any systematic differences. Of 19 variable of interest

on which the two were compared, the sample and the rest of the population differed on three

variables: the sample that was used in the analysis consisted of fewer patients who were

employed (30.1% vs. 35.4%), more oral steroid users (6.3% vs. 4.3%) and more patients who

correctly answered one knowledge item (84% vs. 78%). The differences were not large (the

sample had 5% more unemployed patients, 2% more steroid users and 6% more patients who

answered the knowledge item correctly) but they could have biased the sample. However, the

higher number of unemployed patients was not reflective of other possible underlying

differences such as differences in age, gender or living situation. It appears that the main reason

for this difference was the protocol that screening coordinators followed according to which

follow up calls were done during the business hours of 9am to 5pm when most employed

patients were at work. The difference in percentage of patients who were oral steroid users was

not related to other risk factors for OP and the difference in correctly answering one knowledge

item was not related to knowledge of other OP knowledge items or patient perception of OP

treatment benefits and barriers.

63

5.6 Areas for future research This thesis has identified a number of areas for future research. The review of post-fracture OP

interventions found that health behaviour change theories have not been used to create OP

interventions. Future research should therefore apply these theories to the post-fracture context

and should compare the effectiveness of interventions based on different theories of health

behaviour change.

Future research can test interventions based on the findings from the second part of the thesis

suggesting that patients at risk for not making the link be provided with a clear message

regarding the quality of their bones as “thin” and emphasizing the significance of even one

fragility fracture. Based on this, a question that future research can examine is whether

interventions containing such a message can help patients who are “resistant” to making the OP-

fracture link to actually make this link at follow up. Furthermore, the role that making the OP-

fracture link has, in treatment initiation and persistence, can be assessed. For example, future

research can explore the impact of providing the patient with a navigational tool emphasizing the

OP-fracture link, a personalized OP intervention by a screening coordinator or a brief surgeon-

delivered intervention with a couple of key statements. Long-term impact of such interventions

can be also assessed.

More research that identifies barriers and solutions to more effective OP management from the

patient’s perspective should be conducted. Identifying and exploring other patient-related factors

that could be influential in treatment initiation or making the OP-fracture link should be pursued,

such as the examining of the role of patient’s health literacy, education level, co-morbidities and

cultural and social variables.

64

5.7 Conclusion

Although the effectiveness of OP management depends on fragility fracture patients’ perception

of their bone health and of their fracture as linked to OP, existing OP interventions do not

account for these concepts. This thesis proposes modifications to current OP interventions, and

provides guidance for creation of new, patient-centered interventions. In our current health care

system, improvements in outcomes of OP management cannot be expected unless we help

patients become active participants in the management of their disease.

65

References

1. Osteoporosis: review of the evidence for prevention, diagnosis and treatment and

cost-effectiveness analysis. Executive summary (1998). Osteoporos.Int., 8 Suppl 4, S3-S6.

2. Adachi, J. D., Ioannidis, G., Olszynski, W. P., Brown, J. P., Hanley, D. A.,

Sebaldt, R. J. et al. (2002). The impact of incident vertebral and non-vertebral fractures on health

related quality of life in postmenopausal women. BMC.Musculoskelet.Disord., 3, 11.

3. Allison, P. D. (1999). Logistic Regression Using SAS System: Theory and

Application. Cary, NC: Sas Institute Inc.

4. Ashe, M., Khan, K., Guy, P., Kruse, K., Hughes, K., O'brien, P. et al. (2004).

Wristwatch-distal radial fracture as a marker for osteoporosis investigation: a controlled trial of

patient education and a physician alerting system. J.Hand Ther., 17, 324-328.

5. Astrand, J., Thorngren, K. G., Tagil, M., & Akesson, K. (2008). 3-year follow-up

of 215 fracture patients from a prospective and consecutive osteoporosis screening program.

Fracture patients care! Acta Orthop., 79, 404-409.

6. Astrand, J., Thorngren, K.-G., & Tagil, M. (2006a). One fracture is enough!

Experience with a prospective and consecutive osteoporosis screening program with 239 fracture

patients. Acta Orthopaedica, 77, 3-8.

7. Astrand, J., Thorngren, K.-G., & Tagil, M. (2006b). One fracture is enough!

Experience with a prospective and consecutive osteoporosis screening program with 239 fracture

patients. Acta Orthopaedica, 77, 3-8.

8. Bandura, A. (2004). Health promotion by social cognitive means. Health

Educ.Behav., 31, 143-164.

9. Beaton, D., Jiang, D., Sujic, R., Sale, J., & Bogoch, E. (2009). Factors influencing

the pharmacological management of OP after a fragility fracture: results from the Ontario

Osteoporosis Strategy's Fracture Clinic Screening Program. J Bone Miner.Res. [Suppl 1].

66

10. Beaton, D. E., Bogoch, E., Sujic, R., & Elliot-Gibson, V. (2007). So, how was it?

Patient opinions on osteoporosis education in the fracture clinic setting. Journal of Bone and

Mineral Research 22[S1], s213.

11. Becker, M. H., Radius, S. M., Rosenstock, I. M., Drachman, R. H., Schuberth, K.

C., & Teets, K. C. (1978). Compliance with a medical regimen for asthma: a test of the health

belief model. Public Health Rep., 93, 268-277.

12. Bessette, L., Ste-Marie, L. G., Jean, S., Davison, K. S., Beaulieu, M., Baranci, M.

et al. (2008a). Recognizing osteoporosis and its consequences in Quebec (ROCQ): background,

rationale, and methods of an anti-fracture patient health-management programme.

Contemp.Clin.Trials, 29, 194-210.

13. Bessette, L., Ste-Marie, L.-G., Jean, S., Shawn, D. K., Beaulieu, M., Baranci, M.

et al. (2008b). Recognizing osteoporosis and its consequences in Quebec (ROCQ): Background,

rationale, and methods of an anti-fracture patient health-management programme. Contemporary

Clinical Trials, 29, 194-210.

14. Bierman AS, e. (2010). Project for an Ontario Women's Health Evidence-based

Report (Rep. No. 2).

15. Black, D. M. & Cooper, C. (2000). Epidemiology of fractures and assessment of

fracture risk. Clin.Lab Med., 20, 439-453.

16. Bliuc, D., Eisman, J. A., & Center, J. R. (2006). A randomized study of two

different information-based interventions on the management of osteoporosis in minimal and

moderate trauma fractures. Osteoporos.Int., 17, 1309-1317.

17. Blonk, M. C., Erdtsieck, R. J., Wernekinck, M. G. A., & Schoon, E. J. (2007).

The fracture and osteoporosis clinic: 1-year results and 3-month compliance. Bone, 40, 1643-

1649.

18. Boden, S. D., Einhorn, T. A., Morgan, T. S., Tosi, L. L., & Weinstein, J. N.

(2005). An AOA critical issue. The future of the orthopaedic surgeon-proceduralist or keeper of

the musculoskeletal system? J Bone Joint Surg Am., 87, 2812-2821.

67

19. Bogoch, E. R., Elliot-Gibson, V., Beaton, D. E., Jamal, S. A., Josse, R. G., &

Murray, T. M. (2006). Effective initiation of osteoporosis diagnosis and treatment for patients

with a fragility fracture in an orthopaedic environment. J.Bone Joint Surg.Am., 88, 25-34.

20. Boothroyd, R. I. & Fisher, E. B. (2010). Peers for progress: promoting peer

support for health around the world. Fam.Pract., 27 Suppl 1, i62-i68.

21. Brenneman, S. K., Blau, E. M., Chen, Y., & Abbott, T. A., III (2002). Validation

of a patient questionnaire, Osteoporosis and You, designed to assess osteoperosis-related

attitudes, knowledge and behaviour. J.Bone Miner.Res., 17, S466.

22. Brown, J. P., Fortier, M., Frame, H., Lalonde, A., Papaioannou, A., Senikas, V. et

al. (2006). Canadian consensus conference on osteoporosis, 2006 update.

J.Obstet.Gynaecol.Can., 28, S95-S112.

23. Brown, J. P. & Josse, R. G. (2002). 2002 clinical practice guidelines for the

diagnosis and management of osteoporosis in Canada. CMAJ., 167, S1-34.

24. Burgener, M., Arnold, M., Katz, J. N., Polinski, J. M., Cabral, D., Avorn, J. et al.

(2005). Older adults' knowledge and beliefs about osteoporosis: results of semistructured

interviews used for the development of educational materials. J.Rheumatol., 32, 673-677.

25. Cadarette, S. M., Beaton, D. E., & Hawker, G. A. (2004). Osteoporosis health

belief scale: minor changes were required after telephone administration among women.

J.Clin.Epidemiol., 57, 154-166.

26. Cadarette, S. M., Gignac, M. A., Beaton, D. E., Jaglal, S. B., & Hawker, G. A.

(2007). Psychometric properties of the "Osteoporosis and You" questionnaire: osteoporosis

knowledge deficits among older community-dwelling women. Osteoporos.Int., 18, 981-989.

27. Cadarette, S. M., Gignac, M. A., Jaglal, S. B., Beaton, D. E., & Hawker, G. A.

(2009). Measuring patient perceptions about osteoporosis pharmacotherapy. BMC.Res.Notes, 2,

133.

28. Center, J. R., Bliuc, D., Nguyen, T. V., & Eisman, J. A. (2007). Risk of

subsequent fracture after low-trauma fracture in men and women. JAMA, 297, 387-394.

68

29. Che, M., Ettinger, B., Liang, J., Pressman, A. R., & Johnston, J. (2006).

Outcomes of a disease-management program for patients with recent osteoporotic fracture.

Osteoporosis International, 17, 847-854.

30. Chevalley, T., Hoffmeyer, P., Bonjour, J.-P., & Rizzoli, R. (2002). An

osteoporosis clinical pathway for the medical management of patients with low-trauma fracture.

Osteoporosis International, 13, 450-455.

31. Cline, R. R., Farley, J. F., Hansen, R. A., & Schommer, J. C. (2005). Osteoporosis

beliefs and antiresorptive medication use. Maturitas, 50, 196-208.

32. Clinical Effectiveness and Evaluation Unit.London, R. C. o. P. (2007). National

Clinical Audit of falls and bone health in older people.

33. Connell, F. A., Diehr, P., & Hart, L. G. (1987). The use of large data bases in

health care studies. Annu.Rev.Public Health, 8, 51-74.

34. Cranney, A., Guyatt, G., Griffith, L., Wells, G., Tugwell, P., & Rosen, C. (2002).

Meta-analyses of therapies for postmenopausal osteoporosis. IX: Summary of meta-analyses of

therapies for postmenopausal osteoporosis. Endocr.Rev., 23, 570-578.

35. Cuddihy, M.-T., Amadio, P. C., Gabriel, S. E., Pankratz, V. S., Kurland, R. L., &

Melton III, L. J. (2004). A prospective clinical practice intervention to improve osteoporosis

management following distal forearm fracture. Osteoporosis International, 15, 695-700.

36. Davis, J. C., Ashe, M. C., Guy, P., & Khan, K. M. (2006). Undertreatment after

hip fracture: a retrospective study of osteoporosis overlooked. J.Am.Geriatr.Soc., 54, 1019-1020.

37. Davis, J. C., Guy, P., Ashe, M. C., Liu-Ambrose, T., & Khan, K. (2007).

HipWatch: osteoporosis investigation and treatment after a hip fracture: a 6-month randomized

controlled trial. J Gerontol.A Biol.Sci.Med.Sci., 62, 888-891.

38. Deeks, A. P., Zoungas, S. P., & Teede, H. P. (2008). Risk perception in women: a

focus on menopause. [Article]. Menopause, 15, 304-309.

69

39. DH/SC, L. &. C. d. (2009). Falls and fractures: effective interventions in health

and social care.

40. Doheny, M. O., Sedlak, C. A., Estok, P. J., & Zeller, R. (2007). Osteoporosis

knowledge, health beliefs, and DXA T-scores in men and women 50 years of age and older.

Orthop.Nurs., 26, 243-250.

41. Edwards, B. J., Bunta, A. D., Madison, L. D., DeSantis, A., Ramsey-Goldman, R.,

Taft, L. et al. (2005). An osteoporosis and fracture intervention program increases the diagnosis

and treatment for osteoporosis for patients with minimal trauma fractures. Joint Commission

Journal on Quality & Patient Safety, 31, 267-274.

42. Edwards, B. J., Iris, M., Ferkel, E., & Feinglass, J. (2006). Postmenopausal

women with minimal trauma fractures are unapprised of the existence of low bone mass or

osteoporosis. Maturitas, 53, 260-266.

43. Elliot-Gibson, V., Bogoch, E. R., Jamal, S. A., & Beaton, D. E. (2004a). Practice

patterns in the diagnosis and treatment of osteoporosis after a fragility fracture: a systematic

review. [Review] [55 refs]. Osteoporosis International.15(10):767-78.

44. Elliot-Gibson, V., Bogoch, E. R., Jamal, S. A., & Beaton, D. E. (2004b). Practice

patterns in the diagnosis and treatment of osteoporosis after a fragility fracture: a systematic

review. [Review] [55 refs]. Osteoporosis International.15(10):767-78.

45. Elliot-Gibson, V., Bogoch, E. R., Jamal, S. A., & Beaton, D. E. (2004c). Practice

patterns in the diagnosis and treatment of osteoporosis after a fragility fracture: a systematic

review. [Review] [55 refs]. Osteoporosis International.15(10):767-78.

46. Elliot-Gibson, V., Bogoch, E. R., Jamal, S. A., & Beaton, D. E. (2004d). Practice

patterns in the diagnosis and treatment of osteoporosis after a fragility fracture: a systematic

review. [Review] [55 refs]. Osteoporosis International.15(10):767-78.

47. Feldstein, A., Elmer, P. J., Smith, D. H., Herson, M., Orwoll, E., Chen, C. et al.

(2006). Electronic medical record reminder improves osteoporosis management after a fracture:

A randomized, controlled trial. Journal of the American Geriatrics Society, 54, 450-457.

70

48. Feldstein, A. C., Vollmer, W. M., Smith, D. H., Petrik, A., Schneider, J., Glauber,

H. et al. (2007). An outreach program improved osteoporosis management after a fracture.

Journal of the American Geriatrics Society, 55, 1464-1469.

49. Fiandt, K., Pullen, C. H., & Walker, S. N. (1999). Actual and perceived risk for

chronic illness in rural older women. Clinical Excellence for Nurse Practitioners, 3, 105-115.

50. Fishbein M & Ajzen I. (1975). Belief, Attitude, Intention, and Behavior: An

Introduction to Theory and Research. Reading, MA: Addison-Wesley.

51. Fletcher RH, F. S. (2005). Clinical Epidemiology. (4 ed.) Lippincott Williams &

Wilkins.

52. Fraser, M. & McLellan, A. R. (2004). A fracture liaison service for patients with

osteoporotic fractures. Professional Nurse., 286-290.

53. Funnell, M. M. (2010). Peer-based behavioural strategies to improve chronic

disease self-management and clinical outcomes: evidence, logistics, evaluation considerations

and needs for future research. Fam.Pract., 27 Suppl 1, i17-i22.

54. Gallacher, S. J. (2005). Setting up an osteoporosis fracture liaison service:

Background and potential outcomes. Best Practice and Research in Clinical Rheumatology, 19,

1081-1094.

55. Gardner, M. J., Brophy, R. H., Demetrakopoulos, D., Koob, J., Hong, R., Rana,

A. et al. (2005). Interventions to improve osteoporosis treatment following hip fracture. A

prospective, randomized trial. J Bone Joint Surg.Am., 87, 3-7.

56. Gerend, M. A., Erchull, M. J., Aiken, L. S., & Maner, J. K. (2006). Reasons and

risk: factors underlying women's perceptions of susceptibility to osteoporosis. Maturitas, 55,

227-237.

57. Giangregorio, L., Dolovich, L., Cranney, A., Adili, A., Debeer, J., Papaioannou,

A. et al. (2009). Osteoporosis risk perceptions among patients who have sustained a fragility

fracture. Patient.Educ.Couns., 74, 213-220.

71

58. Giangregorio, L., Papaioannou, A., Cranney, A., Zytaruk, N., & Adachi, J. D.

(2006). Fragility fractures and the osteoporosis care gap: an international phenomenon.

Semin.Arthritis Rheum., 35, 293-305.

59. Giangregorio, L., Papaioannou, A., Thabane, L., Debeer, J., Cranney, A.,

Dolovich, L. et al. (2008). Do patients perceive a link between a fragility fracture and

osteoporosis? BMC.Musculoskelet.Disord., 9, 38.

60. Glanz, K., Rimer, B. K., & Lewis, F. M. (2002). Theory, Research and Practice.

In Health Behavior and Health Education ( San Francisco: Wiley & Sons.

61. Goeree, R. & O'Brien B, P. D. C. L. F. M. A. J. (1996). An assessment of the

burden of illness due to osteoporosis in Canada. J Obstet Gynaecol Can, 18(suppl), 15-24.

62. Gold, D. T. (1996). The clinical impact of vertebral fractures: quality of life in

women with osteoporosis. Bone, 18, 185S-189S.

63. Gold, D. T., Alexander, I. M., & Ettinger, M. P. (2006). How can osteoporosis

patients benefit more from their therapy? Adherence issues with bisphosphonate therapy. The

Annals of Pharmacotherapy, 40, 1143-1150.

64. Gold, D. T. & Silverman, S. L. (2004). Osteoporosis self-management: Choices

For Better Bone Health. South.Med.J., 97, 551-554.

65. Guideline Development Group for the Royal College of Physicians (2010).

Glucocorticoid-induced osteoporosis London: Royal College of Physicians.

66. Harrell, F. (2001). Regression Modeling Strategies with Applications to Linear

Models, Logistic Regression, and Survival Analysis. Springer-Verlag.

67. Harrington, J. T., Barash, H. L., Day, S., & Lease, J. (2005). Redesigning the care

of fragility fracture patients to improve osteoporosis management: A health care improvement

project. Arthritis Care and Research, 53, 198-204.

72

68. Harrington, J. T. & Lease, J. (2007). Osteoporosis disease management for

fragility fracture patients: new understandings based on three years' experience with an

osteoporosis care service. Arthritis Rheum., 57, 1502-1506.

69. Hawker, G., Ridout, R., Ricupero, M., Jaglal, S., & Bogoch, E. (2003). The

impact of a simple fracture clinic intervention in improving the diagnosis and treatment of

osteoporosis in fragility fracture patients. Osteoporosis International, 14, 171-178.

70. Hawker, G. A. (1996). The epidemiology of osteoporosis. J.Rheumatol.Suppl, 45,

2-5.

71. Ho, C., Cranney, A., & Campbell, A. (2006). Measuring the impact of pharmacist

intervention: Results of patient education about osteoporosis after fragility fracture. Canadian

Journal of Hospital Pharmacy, 59, 184-193.

72. Hosmer, D. W. & Lemeshow, S. (2000). Applied Logistic Regression. (2 ed.) New

York: John Wiley & Sons, Inc.

73. Hsieh, C. H., Wang, C. Y., McCubbin, M., Zhang, S., & Inouye, J. (2008).

Factors influencing osteoporosis preventive behaviours: testing a path model. J.Adv.Nurs., 62,

336-345.

74. Jaglal, S. B., Hawker, G., Bansod, V., Salbach, N. M., Zwarenstein, M., Carroll, J.

et al. (2009). A demonstration project of a multi-component educational intervention to improve

integrated post-fracture osteoporosis care in five rural communities in Ontario, Canada.

Osteoporos.Int., 20, 265-274.

75. Jaglal, S. B., Hawker, G., Cameron, C., Canavan, J., Beaton, D., Bogoch, E. et al.

(2010). The Ontario Osteoporosis Strategy: implementation of a population-based osteoporosis

action plan in Canada. Osteoporos.Int., 21, 903-908.

76. Jaglal, S. B., Weller, I., Mamdani, M., Hawker, G., Kreder, H., Jaakkimainen, L.

et al. (2005). Population trends in BMD testing, treatment, and hip and wrist fracture rates: are

the hip fracture projections wrong? J Bone Miner.Res., 20, 898-905.

73

77. Janz, N. K. & Becker, M. H. (1984). The Health Belief Model: a decade later.

Health Educ.Q., 11, 1-47.

78. Jensen, A. L. & Harder, I. (2004). The osteoporotic pain experience.

Osteoporos.Int., 15, 204-208.

79. Johnson, S. L., Petkov, V. I., Williams, M. I., Via, P. S., & Adler, R. A. (2005).

Improving osteoporosis management in patients with fractures. Osteoporosis International, 16,

1079-1085.

80. Kanis, J. A., Burlet, N., Cooper, C., Delmas, P. D., Reginster, J. Y., Borgstrom, F.

et al. (2008). European guidance for the diagnosis and management of osteoporosis in

postmenopausal women. Osteoporos.Int., 19, 399-428.

81. Kanis, J. A. & Johnell, O. (2005). Requirements for DXA for the management of

osteoporosis in Europe. Osteoporos.Int., 16, 229-238.

82. Kanis, J. A., Johnell, O., De, L. C., Johansson, H., Oden, A., Delmas, P. et al.

(2004). A meta-analysis of previous fracture and subsequent fracture risk. Bone, 35, 375-382.

83. Kaufman, J. D., Bolander, M. E., Bunta, A. D., Edwards, B. J., Fitzpatrick, L. A.,

& Simonelli, C. (2003). Barriers and solutions to osteoporosis care in patients with a hip fracture.

J.Bone Joint Surg.Am., 85-A, 1837-1843.

84. Kim, K. K., Horan, M. L., Gendler, P., & Patel, M. K. (1991). Development and

evaluation of the Osteoporosis Health Belief Scale. Res.Nurs.Health, 14, 155-163.

85. Klotzbuecher, C. M., Ross, P. D., Landsman, P. B., Abbott, T. A., III, & Berger,

M. (2000). Patients with prior fractures have an increased risk of future fractures: a summary of

the literature and statistical synthesis. J.Bone Miner.Res., 15, 721-739.

86. Kmetic, A., Joseph, L., Berger, C., & Tenenhouse, A. (2002). Multiple imputation

to account for missing data in a survey: estimating the prevalence of osteoporosis. Epidemiology,

13, 437-444.

74

87. Kuo, I., Ong, C., Simmons, L., Bliuc, D., Eisman, J., & Center, J. (2007).

Successful direct intervention for osteoporosis in patients with minimal trauma fractures.

Osteoporosis International, 18, 1633-1639.

88. Langridge, C. R., McQuillian, C., Watson, W. S., Walker, B., Mitchell, L., &

Gallacher, S. J. (2007). Refracture following fracture liaison service assessment illustrates the

requirement for integrated falls and fracture services. Calcif.Tissue Int., 81, 85-91.

89. Laslett, L. L., Whitham, J. N., Gibb, C., Gill, T. K., Pink, J. A., & McNeil, J. D.

(2007). Improving diagnosis and treatment of osteoporosis: Evaluation of a clinical pathway for

low trauma fractures. Archives of Osteoporosis, 2, 1-6.

90. Levasseur, R., Sabatier, J.-P., Guilcher, C., Guaydier-Souquieres, G., Costentin-

Pignol, V., Jean-Jacques, P.-Y. et al. (2007). Medical management of patients over 50 years

admitted to orthopedic surgery for low-energy fracture. Joint Bone Spine, 74, 160-165.

91. Levinson, M. R. & Clay, F. J. (2009). Barriers to the implementation of evidence

in osteoporosis treatment in hip fracture. Intern.Med.J., 39, 199-202.

92. Lewiecki, E. M. & Watts, N. B. (2009). New guidelines for the prevention and

treatment of osteoporosis. South.Med.J, 102, 175-179.

93. Lindsay, R., Silverman, S. L., Cooper, C., Hanley, D. A., Barton, I., Broy, S. B. et

al. (2001). Risk of new vertebral fracture in the year following a fracture. JAMA, 285, 320-323.

94. Lorig, K. R., Sobel, D. S., Stewart, A. L., Brown, B. W., Jr., Bandura, A., Ritter,

P. et al. (1999). Evidence suggesting that a chronic disease self-management program can

improve health status while reducing hospitalization: a randomized trial. Med.Care, 37, 5-14.

95. Lorrain, J., Paiement, G., Chevrier, N., Lalumiere, G., Laflamme, G. H., Caron, P.

et al. (2003). Population demographics and socioeconomic impact of osteoporotic fractures in

Canada. Menopause., 10, 228-234.

96. MacLean, C., Newberry, S., Maglione, M., McMahon, M., Ranganath, V.,

Suttorp, M. et al. (2008). Systematic review: comparative effectiveness of treatments to prevent

75

fractures in men and women with low bone density or osteoporosis. Ann.Intern.Med., 148, 197-

213.

97. Majumdar, S. R., Beaupre, L. A., Harley, C. H., Hanley, D. A., Lier, D. A., Juby,

A. G. et al. (2007). Use of a case manager to improve osteoporosis treatment after a hip fracture:

Results of a randomized controlled trial. Archives of Internal Medicine, 167, 2110-2115.

98. Majumdar, S. R., Johnson, J. A., McAlister, F. A., Bellerose, D., Russell, A. S.,

Hanley, D. A. et al. (2008). Multifaceted intervention to improve diagnosis and treatment of

osteoporosis in patients with recent wrist fracture: a randomized controlled trial. CMAJ., 178,

569-575.

99. Majumdar, S. R., Rowe, B. H., Folk, D., Johnson, J. A., Holroyd, B. H., Morrish,

D. W. et al. (2004). A controlled trial to increase detection and treatment of osteoporosis in older

patients with a wrist fracture. Annals of Internal Medicine, 141, 366-54.

100. Mauck, K. F., Cuddihy, M. T., Trousdale, R. T., Pond, G. R., Pankratz, V. S., &

Melton, L. J., III (2002). The decision to accept treatment for osteoporosis following hip

fracture: exploring the woman's perspective using a stage-of-change model. Osteoporos.Int., 13,

560-564.

101. McBride, C. M., Emmons, K. M., & Lipkus, I. M. (2003). Understanding the

potential of teachable moments: the case of smoking cessation. Health Educ.Res., 18, 156-170.

102. McBride, C. M., Puleo, E., Pollak, K. I., Clipp, E. C., Woolford, S., & Emmons,

K. M. (2008). Understanding the role of cancer worry in creating a "teachable moment" for

multiple risk factor reduction. Soc.Sci.Med., 66, 790-800.

103. McCabe, J. A. (2006). An assignment for building an awareness of the

intersection of health literacy and cultural competence skills. J Med.Libr.Assoc., 94, 458-461.

104. McLellan, A. R., Gallacher, S. J., Fraser, M., & McQuillian, C. (2003). The

fracture liaison service: Success of a program for the evaluation and management of patients

with osteoporotic fracture. Osteoporosis International, 14, 1028-1034.

76

105. Meadows, L. M., Mrkonjic, L., & Lagendyk, L. (2005). Women's perceptions of

future risk after low-energy fractures at midlife. Ann.Fam.Med., 3, 64-69.

106. Meadows, L. M. & Mrkonjic, L. A. (2003). Breaking--bad news: women's

experiences of fractures at midlife. Can.J.Public Health, 94, 427-430.

107. Meadows, L. M., Mrkonjic, L. A., Lagendyk, L. E., & Petersen, K. M. (2004).

After the fall: women's views of fractures in relation to bone health at midlife. Women Health,

39, 47-62.

108. Melton, L. J., III (1995). How many women have osteoporosis now? J.Bone

Miner.Res., 10, 175-177.

109. Monti, P. M., Colby, S. M., Barnett, N. P., Spirito, A., Rohsenow, D. J., Myers,

M. et al. (1999). Brief intervention for harm reduction with alcohol-positive older adolescents in

a hospital emergency department. J.Consult Clin.Psychol., 67, 989-994.

110. Mulligan, K. & Newman, S. (2003). Psychoeducational interventions in rheumatic

diseases: a review of papers published from September 2001 to August 2002.

Curr.Opin.Rheumatol., 15, 156-159.

111. Munro, S., Lewin, S., Swart, T., & Volmink, J. (2007). A review of health

behaviour theories: how useful are these for developing interventions to promote long-term

medication adherence for TB and HIV/AIDS? BMC.Public Health, 7, 104.

113. National Institute for Health and Clinical Excellence. (2008). Alendronate,

etidronate, risedronate, raloxifene, strontium ranelate for the primary prevention of osteoporotic

fragility fractures in postmenopausal women. Final Appraisal Determination. TAG 160.

114. National Library of Medicine (2010). Current bibliographies in medicine: health

literacy. http://www.nlm.nih.gov/pubs/cbm/hliteracy.html

115. National Osteoporosis Foundation (2010). Clinician's Guide to Prevention and

Treatment of Osteoporosis Washington, DC: National Osteoporosis Foundation.

77

116. NIH Consensus Development Conference on Osteoporosis Prevention, D. a. T.

(2000). Osteoporosis Prevention,Diagnosis,and Therapy. NIH Consens.Statement, 17, 1-45.

117. Papaioannou, A., Clarke, J. A., Campbell, G., & Bedard, M. (2000). Assessment

of adherence to renal dosing guidelines in long-term care facilities. J.Am.Geriatr.Soc., 48, 1470-

1473.

118. Papaioannou, A., Watts, N. B., Kendler, D. L., Yuen, C. K., Adachi, J. D., &

Ferko, N. (2002). Diagnosis and management of vertebral fractures in elderly adults. Am.J.Med.,

113, 220-228.

119. Persell, S. D., Dolan, N. C., Friesema, E. M., Thompson, J. A., Kaiser, D., &

Baker, D. W. (2010). Frequency of inappropriate medical exceptions to quality measures.

Ann.Intern.Med., 152, 225-231.

120. Popa, M. A. (2005). Stages of change for osteoporosis preventive behaviors: a

construct validation study. J.Aging Health, 17, 336-350.

121. Prochaska, J. O. & Velicer, W. F. (1997). The transtheoretical model of health

behavior change. Am.J.Health Promot., 12, 38-48.

122. Prochaska, J. O., Velicer, W. F., Rossi, J. S., Goldstein, M. G., Marcus, B. H.,

Rakowski, W. et al. (1994). Stages of change and decisional balance for 12 problem behaviors.

Health Psychol., 13, 39-46.

123. Quintos-Macasa, A. M., Quinet, R., Spady, M., Zakem, J., Davis, W., Menon, Y.

et al. (2007). Implementation of a mandatory rheumatology osteoporosis consultation in patients

with low-impact hip fracture. J.Clin.Rheumatol., 13, 70-72.

124. Rosenstock, I. M., Strecher, V. J., & Becker, M. H. (1988). Social learning theory

and the Health Belief Model. Health Educ.Q., 15, 175-183.

125. Ryan, R. M., Patrick, H., Deci, E. L., & Williams, G. C. (2008). Facilitating

health behaviour change and its maintenance: Interventions based on Self-Determination Theory.

The European Health Psychologist, 10.

78

126. Schmid, L., Henzen, C., Schlumpf, U., & Babst, R. (2004). Improving secondary

prevention in fragility fracture patients: the impact of a simple clinical information procedure.

Journal of Applied Research.2004; 4, 570-575.

127. Simonelli, C., Killeen, K., Mehle, S., & Swanson, L. (2002). Barriers to

osteoporosis identification and treatment among primary care physicians and orthopedic

surgeons. Mayo Clin.Proc., 77, 334-338.

128. Skedros, J. G. (2004). The orthopaedic surgeon's role in diagnosing and treating

patients with osteoporotic fractures: standing discharge orders may be the solution for timely

medical care. Osteoporos.Int., 15, 405-410.

129. Tenenhouse, A., Joseph, L., Kreiger, N., Poliquin, S., Murray, T. M., Blondeau,

L. et al. (2000). Estimation of the prevalence of low bone density in Canadian women and men

using a population-specific DXA reference standard: the Canadian Multicentre Osteoporosis

Study (CaMos). Osteoporos.Int., 11, 897-904.

130. Tosi, L. L., Gliklich, R., Kannan, K., & Koval, K. J. (2008). The American

Orthopaedic Association's "own the bone" initiative to prevent secondary fractures. J Bone Joint

Surg.Am., 90, 163-173.

131. Tugwell, P. S., Santesso, N. A., O'Connor, A. M., & Wilson, A. J. (2007).

Knowledge translation for effective consumers. Phys.Ther., 87, 1728-1738.

132. U.S.Department of Health and Human Services (2004). The 2004 Surgeon

General's Report on Bone Health and Osteoporosis: What It Means To You U.S. Department of

Health and Human Services, Office of the Surgeon General.

133. Vaile, J., Sullivan, L., Bennett, C., & Bleasel, J. (2007). First Fracture Project:

addressing the osteoporosis care gap. Intern.Med.J., 37, 717-720.

134. van Staa, T. P., Dennison, E. M., Leufkens, H. G., & Cooper, C. (2001).

Epidemiology of fractures in England and Wales. Bone, 29, 517-522.

135. Wallace, L. S. (2002). Osteoporosis prevention in college women: application of

the expanded health belief model. Am.J.Health Behav., 26, 163-172.

79

136. Wichowski, H. C. & Kubsch, S. M. (1997). The relationship of self-perception of

illness and compliance with health care regimens. J.Adv.Nurs., 25, 548-553.

137. Wiktorowicz, M. E., Goeree, R., Papaioannou, A., Adachi, J. D., &

Papadimitropoulos, E. (2001). Economic implications of hip fracture: health service use,

institutional care and cost in Canada. Osteoporos.Int., 12, 271-278.

80

Appendix I

Description of the studies included in the literature review of patient-centered post-fracture osteoporosis interventions in the

context of theories of behaviour change

First author year

Study design or program type

Fracture type and patients’ age

Theoretical framework specified

Elements of intervention

Ashe 2004

Controlled trial

Low trauma wrist fx 50+

No Usual care (BMD offered, specialist referral, discussion of meds, exercise prescription and lifestyle modification) vs. intervention (in addition to usual care): letter to patient explaining that she or he has suffered a low trauma fx diagnostic of OP, letter for GP for patient to take, FU telephone call to patient to remind of visit with GP, faxed letter to GP

Astrand 2006 Astrand 2008

Screening program in orthopaedic department

Fragility fxs 50-75

No, only premise that patients can be trusted to seek appropriate medical advice.

Patients given: one letter for GP, another one for themselves, a copy of the BMD and questionnaire (on risk factors) that they completed. Diagnosed patients referred to GP.

Bessette 2008

Randomized controlled trial/ ongoing provincial

Women 50+ post-fx

No All patients told about the link between fragility fx and another fx. Documentation group received written OP information emphasizing the link between

Description of the studies included in the review Abbreviations: PCP=primary care practitioner GP=general practitioner OP=osteoporosis fx(s)=fracture(s) tx=treatment BMD=bone mineral density FU=follow up

81

OP health management program

OP and fragility fx, patients’ GPs received written information. Intervention group received documentation group material and detailed video about link between OP and fx.

Bliuc 2006

Randomized trial

Low or minimal trauma patients

No Information-based interventions. Both control and intervention groups received personalized letter with participant's risk factors and recommendation for FU with a PCP. Intervention group also included an offer for a free BMD test.

Blonk 2007

Screening program: Fracture and OP clinic

Low trauma 50+

No BMD arranged, results explained, questionnaire including risk factors and life habits explained, diagnosis and tx strategy explained by a nurse

Bogoch 2006

Screening program: OP exemplary care program

Fragility fxs Women 40+ Men 50+

No Individual counselling assessing risk factors, questionnaire, BMD arranged, visit to a specialist booked, summary sent to GP

Che 2006

Pilot study of an intervention (disease-management) program

Osteoporotic fxs 55+

No Lab tests and BMD done. Patients received written information and counselling over the phone: fx risk score, link between OP and fx reviewed. Meds offered. Summary sent to both PCP and patient.

Chevalley 2002

Screening program, OP Clinical pathway

Low trauma any age

No Interactive educational program for patients and families on the management of OP and lifestyle, physical therapy and nutrition (done by multidisciplinary team). BMD and tx suggestions sent to PCP and orthopaedic surgeon.

Cuddihy 2004

Clinical practice intervention

Moderate impact distal forearm fx 45+

No Patients provided with written information and filled out OP quality of life questionnaire. BMD test booked. Appointment

82

with GP facilitated. Davis 2007

Randomized controlled trial: Patient Empowerment and Physician Alert

Hip fxs Women 60+

No. The study tested an intervention involving patient empowerment

Intervention group: patients given written information, asked to see a GP and take a letter to GP.

Edwards 2005

Screening program: OP and fx intervention program

Minimal trauma fx 40+

No Patients given printed material on OP, diet, home safety and educational consult, lifestyle modification, BMD recommended, tx initiated if needed (recommendations mailed to PCP at first, then hospital resources mobilized).

Feldstein 2006

Randomized controlled trial

Women 50-89

No Usual care or PCP reminder or PCP reminder plus education (patient were mailed written letter with OP information about their risk, guidelines for management and general OP information. Physicians received the same material and an electronic record reminder).

Feldstein 2007

Outreach program evaluated – retrospective cohort

Women 67+

No Patients received counselling regarding their risk for OP, fx and lifestyle. BMD results explained and meds initiated. PCPs sent an electronic record reminder with guidelines and suggested tx.

Fraser 2004 Gallacher 2005 Langridge 2007 McLellan 2003

Fracture Liaison Service

Fragility fx 50+

No Patients received an educational consult with a nurse, BMD arranged, tx recommended if needed. Assessment letter sent to GP.

Gardner 2005

Randomized controlled

Low trauma hospitalized

No Control group: given a 2 page pamphlet on fall prevention, not

83

trial hip fx patients

going into detail about OP. Contacted after 6 months to see if OP was addressed by DXA or tx. Study group: coordinator discussed OP, future fx, available tx. Patients given 5 questions to ask their GP. Contacted after 6 weeks to remind about importance of FU and management of OP.

Harrington 2005 Harrington 2007

Clinical improvement project: Osteoporosis Care Service

Fragility fxs 50+

No Explained the service to patients and family. Provided education and coordination of care, patient advised to contact PCP for OP management or referred to OP specialty clinic, patient complete a clinical questionnaire, assigned a nurse manager to answer questions and respond to problems.

Hawker 2003

Pre-post intervention design – pilot test of a simple intervention

Fragility fxs 40+

No Patients informed by orthopaedic that fx likely OP related and patients at risk for OP and fxs. Patients encouraged to FU with GP and given a standardized letter.

Ho 2006

Non- randomized prospective evaluation of a pilot intervention

Fragility fxs 45+

No Patients interviewed by pharmacy resident to identify barriers, patients also received an educational intervention (presentation on OP/falls/diet etc and written information), completed questionnaire on OP knowledge, encouraged to speak to physiotherapist and see a GP.

Johnson 2005

Prospective study - simple intervention

Fx patients Veterans (95% males)

No Patients were informed about the importance of diagnosing and treating OP, risk factors (also given written information). BMD test offered.. Report sent to PCP. Supplements recommended if needed.

84

Kuo 2007 Prospective cohort

Minimal trauma Adults

No Control group: information only. Intervention group: Face to face counselling re OP, patient-specific risk factors and lifestyle change. Results of BMD discussed, supplements and antiresorptive medications discussed and recommended (not prescribed). Information sent to GP and patients referred for BMD.

Laslett 2007

Clinical pathway

Admitted fragility fxs 45+

No Those patients recommended/started therapy also received individual counselling as well as written information about OP, letter to GP and resource information. No investigation of OP was undertaken – left to GP.

Levasseur 2007

Ortho department intervention

Low energy fx admitted to orthopaedic dept 50+

No During interview with rheumatologist, lifestyle changes and tx discussed. BMD performed. Letter to GP sent for those patients who were given a prescription at discharge.

Majumdar 2004

Non-randomized controlled trial

Wrist fx patients 50+

No Faxed physician reminders along with guidelines endorsed by local opinion leaders. Patient education: written message encouraging patients to seek further information from GP reinforced by a brief telephone counselling session.

Majumdar 2007

Randomized controlled trial

Hip fx patients 50+

No Intervention group consisted of usual care and counselling about the importance of BMD and tx in reducing the risk of another fx. BMD arranged and results discussed. Treatment facilitated. Summary sent to GP.

Majumdar 2008

Randomized controlled trial

Wrist fx patients 50+

No Control (pamphlet only) vs. Intervention: patients given the same pamphlet and counselling

85

reinforcing it. GP’s were mailed reminders and local opinion leader-endorsed OP tx guidelines.

Quintos-Macasa 2007

Retrospective cohort to examine rheumatology service

Low impact hip fxs

No Patients were prescribed medication after reviewing their records and scheduling a BMD if necessary. Rheumatology service communicated with PCPs.

Schmid 2004

Prospective intervention “OsteoCare project”

Low trauma 50-85 years old

No

Phase 1: Surgeons were asked to inform patients about OP possibly causing fx and give them written material including recommendation to see their GP for further evaluation of OP. Phase 2: GPs were informed about the project and asked to initiate diagnostic OP procedures.

Skedros 2004

Program aimed at surgeons (and indirectly at patients)

Low trauma 50+

No Ortho surgeons recruited to tell patients that they may have OP, were at risk of another fx and should see their GP. Ortho surgeons sent 2 letters to patients GP’s.

Tosi 2008 Evaluation of the “Own the Bone” tool /pilot intervention

Low energy fxs Any age

No Prompts to provide counselling to patients re OP, fall prevention, exercise and tx. Letter sent to GP.

Vaile 2007

“First Fracture Project” evaluation

Low trauma 50+

No, only that the intervention was aimed at patients with ‘fresh’ fx pain

Patients were educated about diet, exercise, risk factor reduction and falls prevention; BMD done, tx started if needed, letter sent to GP. Follow up at 1 month to encourage compliance with meds then a year after to repeat BMD and check for Vit D.

86

Appendix II

Fracture Clinic Screening Program’s patient educational materials and Program Information Letter

87

88

Appendix III

Construct list with description of all variables collected as part of the Fracture Clinic Screening Program

Data element (Number in

survey)

Field Description

Patient ID Numeric Provide link to de-identified information Facility ID Numeric Linked to patient ID DISCLAIMER 1 item Coordinator has read the disclaimer to the patient Health Card Number Numeric Only if the patient consents to his/her health card

number to be collected Agreed to health card number Yes/No Indicator that subject has agreed to link with health

data. Health card number is with identifiable data. Consent signed/stored Yes/No Indicator for central database that consent was

signed and is on site at clinic Gender Alpha Age Years Year of birth Numeric Last Name Alpha On contact sheet and consent form First Name Alpha On contact sheet and consent form Address On contact sheet Telephone Number(s) On contact sheet

Full date of birth Numeric Date completed Coordinator name

Marital Status 1 item 6 responses (choose 1)

Never married (single); Common Law; Married; Separated; Divorced; Widowed

Language 1 item 3 responses (choose 1)

English; French; Other (specify)

First 3 digits of postal code 1 item e.g. M5R

Date of fracture Numeric On baseline questionnaire Living Arrangement 1 question Where does this patient live?

Employment at time of fracture and 12 weeks post-fracture

1 item Yes/No

Baseline: At the time of this patient’s fracture, was he/she working full or part-time for pay? Follow up: Were you employed full time or part time at the time of your fracture?

89

Not employed at time of fracture: reason

1 item 7 responses (choose 1)

Homemaker; Unable to find work; Not working by choice; Volunteering; Retired; Student; Not working due to health conditions or disability.

Not employed at follow up (if employed at baseline)

1 item 2 responses

Not working due to fracture Other

Able to complete full or partial documentation? Yes/No If NO, why? Select one: Language; Mental/Physical barrier; Patient declined Other (specify).

Ability to complete Baseline Documentation

4 questions

Is there someone else able to complete partial documentation? Yes/No If YES, who: Patient; family member; friend; translator; chart

Individual who completes documentation

1 question 5 responses (choose 1)

Patient; family member; friend; translator; chart

Fracture location

1 question, 10 responses (check all that apply)

Provided by x-ray results; surgeon; and/or patient: Wrist (distal radius), Elbow, Shoulder, Hip, Vertebrae, Pelvis Tibia/fibula, Femur (not hip), Ankle, Other: specify

Fracture etiology (fragility fracture defined by WHO [51]

1 question 3 responses (choose 1)

Provided by patient, family member or surgeon Low trauma; Moderate/high trauma; Unknown

History previous fragility fracture

1 question Yes/No

Since you were 40, have you broken any other bones?

Trouble getting out of chair or problem with walking

1 question Yes/No

Do you have trouble getting out of a chair or feel unsteady when you walk?

Maternal fracture 1 question Yes/No

Did this patient’s biological mother ever have a hip fracture after the age of 40 years? Other bones?

History of falls 1 question Yes/No

Have you fallen in the past year besides the fall that may have led to your current fracture?

Weight Kilograms and pounds ______ lbs or _____ kgs

Steroid Use 1 question Have you ever taken steroids, such as prednisone, by mouth for more than 3 months at a time?

Alcohol Use 1 question On average, do you drink more than 2 alcoholic beverages a day?

Diagnosis of rheumatoid arthritis 1 question Has a physician specialist, such as rheumatologist,

ever told this patient that they have…? Did this physician prescribe medications…? Medications for

rheumatoid arthritis 2 questions What medications did their physician prescribe…?

90

Smoking 1 question: 3 answers

Does this patient currently smoke cigarettes? Yes, No, Refused Baseline: Prior to having this broken bone, was this patient ever told by a physician that they have OP or osteopenia? Diagnosis of

osteoporosis pre and post fracture by a physician

1 question Yes/No

Follow up: In the past 12 weeks has your family physician….? Predictive of treatment * not true diagnosis, but more perceived diagnosis

Physician management of osteoporosis

1 question 8 responses

Family physician (or general practitioner); Rheumatologist; Endocrinologist, Gynecologist; Internist; Orthopaedic Surgeon; No one, Other (specify)

Physician recommend supplements/ prescribe medication for osteoporosis/ osteopenia

1 question Yes/no

Did this patient’s physician recommend supplements or…?

Supplements recommended/ medications prescribed for osteoporosis

1 question 14 responses

Calcium Supplements, Vitamin D, Didrocal ®, Fosamax ®, Fosavance®, Actonel®, Calcimar ® Miacalcin NS®, HRT, Aredia ®, Forteo®, Evista®, Ipriflavone, Patient in clinical trial, Other: specify

Adherence with medications

1 question Yes/no Patient currently taking medications prescribed?

Reasons for non-adherence with medications prescribed

1 question 10 responses

Medication adherence responses(follow up variable names in brackets): Side effects of the medications; Cost (too expensive); On too many medications already; Forgot to take them; I have decided not to take any medications; I have not made up my mind whether to take the medications or not; I have not started taking them yet I cannot take the medications because of: kidney disease, abnormalities of the esophagus, cannot swallow pills who; Other (specify).

Completion of bone mineral density (DXA) in past year

1 question Yes/No/ Don’t know

Knowledge of bone mineral density results (personal communication, J.

1 question 5 responses

BMD testing responses: (modified from Brown, 2006) Yes, the results show I have osteoporosis; Yes, the results show I have osteopenia;

91

Brown, 2006) Yes, the results show I have low bone density, but I do not know if I have osteopenia or osteoporosis; Yes, the results show I have normal bone density or high bone density; No, I do not know the results of my BMD test.

Osteoporosis Readiness for change (Mauck, 2002) [53]

1 question 8 responses

Responses: Never heard of OP; Aware, never thought seriously; Considered, but decided against; Currently considering treatment; Decided to use treatment but not yet started Recently started on treatment (modified response) <6 months; Been on treatment > 6 months; Refused

Patient opinion of bone quality (personal communication, G. Hawker, 2003)

1 question 5 responses (chose one)

OP Knowledge and Beliefs responses: My bone are normal; My bones are normal for someone my age; My bones may be thin; My bone are definitely thin (modified response); I do not know,

Did osteoporosis cause you to break your bone?

1 question Yes/ No/ Don’t know

(Personal communication, G. Hawker, 2003)

Osteoporosis Knowledge (Cadarette, 2006) [54]

4 questions Response: Strongly agree to Strongly disagree

Questions: There is no way to prevent osteoporosis Bones cannot be rebuilt once they are thin from osteoporosis If a person has osteoporosis, something as simple as lifting a bag of groceries can break a bone The health problems caused by osteoporosis can be life threatening

Osteoporosis drug treatment Caderette, 2006 [54]

7 questions Responses: Strongly agree to Strongly disagree

Questions: Drug treatments can help to build strong bones You would feel good about taking drug treatments to treat osteoporosis (modified from original) Drug treatments can cut down the chances of broken bones You would consider taking drug treatments to prevent broken bones If your doctor advised you to, you would take drug treatments to prevent broken bones You are taking too many medications You have a stomach problem that limit your ability to take drug treatment

Patient has a family physician?

2 questions Yes/No

Documentation of name, address, phone, fax of family physician Referral to College of Physicians and Surgeons if patient does not have a family physician

Recommendations from Osteoporosis Screening Coordinator

Check list

Standardized letter to family physician Discuss completing bone mineral density (DXA) test with family physician Discuss calcium and vitamin D supplements with pharmacist

92

Provide patient with information letter on program

Visit date Please specify the date of the visit Follow up with physician 1 question Did you see a physician for assessment of OP…?

Type of physician 1 question 9 responses What type of physician did you see?

Reasons for not following up with family physician for osteoporosis assessment

1 item 7 responses (choose 1)

Did not obtain a family physician My family physician has referred me to a specialist for assessment of bone health. Forgot/missed an appointment with my family physician Did not have time to see my family physician Appointment with family physician is coming up Did not want to follow up with family physician Other (please write):____________

OSC suggestion "I suggest you follow up with your family physician for further assessment…" Can I call you…"

DXA Testing Completed in past 12 or 24 weeks

Item Yes/No/Don’t Know

Bone mineral density (DXA) test completed (self-report)

Reasons for not completing a DXA test in 12 or 24 weeks

1 item 9 responses

Not recommended to have a bone mineral density test I do not have a family a family physician, therefore did not get a referral for this test I did not go to see my GP My physician said I did not need a bone mineral density test My appointment is coming up I missed/forgot my appointment I did not want to have a bone mineral density test Other (please write) Refused Did you miss work due to fracture? Time off work due

to fracture 2 questions How much time did you have to take off from your job (weeks)?

Quality of information and care

1 question What is your overall opinion of the quality of care…?

Information received on OP 2 questions In the past 12 weeks did you call…?

General comments General comments for the FU visit

93

Appendix IV

Research ethics approval from the University of Toronto and St. Michael’s hospital

94