Osteoporosis International

, Volume 16, Issue 5, pp 447–455

Quality of life in patients with osteoporosis

Authors

    • Department of EndocrinologyVU University Medical Center
    • Institute for Extramural Medicine (EMGO)VU University Medical Center
  • Natasja M. van Schoor
    • Institute for Extramural Medicine (EMGO)VU University Medical Center
Review

DOI: 10.1007/s00198-004-1762-7

Cite this article as:
Lips, P. & van Schoor, N.M. Osteoporos Int (2005) 16: 447. doi:10.1007/s00198-004-1762-7

Abstract

Complaints regarding, and morbidity of, osteoporosis are caused by fractures which are associated with pain and decrease of physical function, social function, and well-being. These are aspects of quality of life. Health-related quality of life covers physical, mental, and social well-being. Quality of life may be measured for evaluation of treatment effects in clinical trials, for the assessment of the burden of the disease of osteoporosis, and for estimates of the cost-effectiveness of different treatment scenarios in health care policy. Quality of life has been measured in patients with osteoporosis with generic questionnaires such as SF-36 and EQ-5D, which can be used in many diseases, or with one of the six available osteoporotic-specific questionnaires, e.g., Qualeffo-41 or OPAQ. Every questionnaire has to be validated to assess psychometric properties and discrimination power between patients with osteoporosis and control subjects. The value attached to specific health states (utility) can be assessed with some generic instruments or by systematic questioning of the patient, e.g., the time-trade-off method. This results in one value for health status ranging from 0 (death) to 1 (perfect health). Utility values can be used to calculate loss of quality-adjusted life years (QALY). Most data have been obtained in patients with prevalent vertebral fractures. Scores of specific and generic questionnaires showed significant loss of quality of life with prevalent vertebral fractures. In addition, studies with Qualeffo-41 and OPAQ showed a deteriorating quality of life with increasing number of vertebral fractures. Lumbar fractures had more impact on quality of life than thoracic fractures. Incident vertebral fractures were also associated with a decrease of quality of life especially in the physical function domain. This applied to clinical incident vertebral fractures as well as to subclinical fractures to a lesser degree. Loss of quality of life following hip fracture has been documented with generic and osteoporosis-specific questionnaires. A considerable loss was observed in the 1st year with some improvement in the 2nd year, but not to baseline values. Quality of life depended on comorbidity, mobility, activities of daily life (ADL)–independence, and fracture complaints. Utility loss has been observed following hip fracture, especially disabling hip fracture, hip and vertebral fracture combined, or multiple vertebral fractures. Utility following osteoporotic fractures has been valued by patients, the healthy elderly, and panels of experts. The healthy elderly gave the worse quality-of-life scores (lower utility) to various hip fractures than patients with hip fractures themselves. In conclusion, suitable instruments exist for measuring quality of life in patients with osteoporotic fractures. These instruments are useful for clinical trials and for assessment of the burden of disease.

Keywords

OsteoporosisQuality of life

Introduction

Osteoporosis results in fractures after minimal trauma. The morbidity of osteoporosis is mainly caused by fractures of the hip, vertebrae, and distal radius [1]. Hip fractures cause acute pain and loss of function and nearly always lead to hospitalization. Recovery is slow and rehabilitation is often incomplete. Many patients are permanently institutionalized in a nursing home. Vertebral fractures may cause acute pain and loss of function but also may pass without serious symptoms [2, 3, 4]. Vertebral fractures often recur, and the consequent disability increases with the number of fractures. Distal radius fractures also lead to acute pain and loss of function, but functional recovery usually is good or excellent. In addition to pain and disturbance of physical function, a fracture may decrease mobility and social interaction and cause emotional problems [5]. All of these characteristics determine quality of life.

Quality of life covers all aspects of life including health status, environment, financial aspects, and human rights. Health-related quality of life or health status is a subset of quality of life, concerning physical, emotional, and social well-being [6, 7, 8]. In this review, the term quality of life is used for health-related quality of life. Why should we measure quality of life? It can be considered a more or less objective assessment of subjective feelings. Treatment effects in clinical trials can be assessed with quality-of-life questionnaires [4]. Quality of life is often used as a secondary endpoint in addition to radiographic and biochemical variables or events such as fractures. In epidemiologic surveys, quality of life may be assessed to obtain data on the burden of disease, complementary to data on morbidity and health care use, and often in comparison with other diseases. Quality-of-life data can also be used to assess the cost-effectiveness of one treatment in comparison with other treatments or with no treatment at all [9]. Quality of life can either be expressed in arbitrary units (e.g., from 0 to 1, or from 0 to 100] or in quality-adjusted life years (QALYs) to facilitate comparison between treatment groups or various diseases [10].

In this review, different methods to assess quality of life will be discussed, including generic questionnaires and disease-specific questionnaires for osteoporosis. Utility, i.e., the assessment of patient preferences, will be discussed. Subsequently, quality-of-life data obtained in observational studies and clinical trials in patients with vertebral fractures will be reviewed. Finally, the data obtained in patients with distal radius fractures and hip fractures will be discussed.

The assessment of quality of life

Quality-of-life questionnaires can be classified into generic, disease-specific, and study-specific questionnaires. Generic questionnaires pose general questions on health status, and can be used in various diseases. They enable comparison between different diseases. On the other hand, they may contain superfluous questions, as they are not specific for any disease or age group. Examples of generic questionnaires are the Nottingham Health Profile (NHP) [11], the Sickness Impact Profile (SIP) [12], the Short Form 36 of the Medical Outcomes Study (SF-36) [13, 14], and the EuroQol (EQ-5D) [15].

Disease-specific or disease-targeted questionnaires are designed for patients with a specific disease—e.g., depression, myocardial infarction, or osteoporosis. They are meant for use in one disease or group of diseases. As these questionnaires contain more specific questions, they may be less of a burden for the patients, because they may recognize their individual problems (Table 1). A disadvantage is that different diseases cannot be compared [6, 7]. This is the reason that many investigators recommend combinations of generic and specific questionnaires, and many clinical trials contain one of each type. Examples of disease-specific questionnaires are the Geriatric Depression Scale [16], the Inflammatory Bowel Disease Questionnaire, for patients with inflammatory bowel disease [17], and the Qualeffo-41, for patients with osteoporosis [18]. Both generic and disease-specific questionnaires usually consist of several domains—e.g., pain, physical function, mobility, general health, emotions, and fears.
Table 1

Characteristics of generic and disease-specific questionnaires for evaluation of quality of life

Generic

Disease-specific

Suitable for many diseases

Suitable for one disease only

May contain superfluous questions

Less of a burden for the patient

Enables comparison between diseases

Comparison between diseases impossible

Examples: SF-36, Euroqol, NHP, SIP

Examples: Qualeffo-41, OPAQ, OQLQ

Validation of quality-of-life instruments

Each quality-of-life questionnaire has to be thoroughly validated before general use can be recommended. First, the psychometric characteristics of the questionnaires should be evaluated [18]. These include floor and ceiling effects, indicating the percentage of subjects with the lowest or highest possible scores. When one of these is high (e.g., 50%), the questionnaire may not be suitable for the tested group. The convergent and discriminant validity concern the homogeneity of questions in one domain. The convergent validity indicates the percentage of question scores correlating better than 0.40 with the domain score. The discriminant validity indicates the percentage of question scores correlating better with its domain score than with the scores of other domains. Repeatability (reproducibility) is the agreement between subsequent administrations of a questionnaire within some weeks. It is calculated as Cohen’s κ and should be better than 0.60. The internal consistency is usually tested by Cronbach α. The closer the proportion of variance due to error (in relation to real variance) is to zero, the closer Cronbach α is to 1. The discriminant power of a questionnaire between patients with the disease and controls can be assessed with logistic regression analysis to derive an odds ratio, with 95% confidence intervals. In addition, discrimination can be investigated by constructing receiver operating characteristic (ROC) curves [18]. This offers the possibility to compare the discriminant power of questionnaires as expressed in sensitivity and specificity over all possible cut-off values (Fig. 1). The area under the ROC curve may vary between 0.5 (no discrimination) and 1.0 (perfect discrimination).
Fig. 1

Receiver operating characteristic (ROC) curves for individual Qualeffo domains in discriminating performance between patients with vertebral fractures and control subjects (reprinted from Lips P et al., Osteoporos Int 10:150–160, 1999)

Osteoporosis-specific questionnaires

Six questionnaires have been developed for patients with osteoporosis, and other questionnaires have been derived from these. The characteristics of these questionnaires are summarized in Table 2. These questionnaires have been developed along different lines, and they exhibit different characteristics. The Osteoporosis Quality of Life Questionnaire (OQLQ) was developed by identifying items in generic questionnaires, and asking patients, nurses, specialists, physiologists, and rheumatologists about them. In this way, 168 items were identified, which were condensed into 30 questions in several steps following well-defined procedures [19, 20]. The OQLQ is interviewer based and also exists in a shortened version with 10 questions [21].
Table 2

Characteristics of quality-of-life questionnaires specific for osteoporosis

Name

Mode of administration

Number of questions

Domains

References

OQLQ

Interviewer

30

Physical function, ADL, emotional function

[19, 20]

OFDQ

Interviewer

69

General health + back pain, ADL, socialization, depression (CES-D), confidence

[22]

OPTQOL

Interviewer

33

Physical activity, adaptations, fears

[23, 24]

OPAQ

Self-administration

67

Physical function, emotional status, symptoms, social interaction

[25]

Qualeffo-41

Self-administration

41

Pain, physical function, social function, general health perception, mental function

[5, 18]

QUALIOST

Self-administration

23

Physical function, emotional status

[27]

The Osteoporosis Functional Disability Questionnaire (OFDQ) was developed to assess disability in patients with osteoporosis and back pain due to vertebral fractures, including pain indexes, a depression scale, and measurements of functional abilities, social activities, and confidence in treatment [22].

The Osteoporosis-Targeted Quality of Life Questionnaire (OPTQOL) was developed as an instrument for community- or population-based studies to assess the burden of osteoporosis in women living in the community. The items in this questionnaire were identified in the literature and in focus groups. The 299 identified items were reduced to 37 items through ranking by a large group of women and after validation reduced to 33 questions [23, 24].

The Osteoporosis Assessment Questionnaire (OPAQ) was developed from the Arthritis Impact Measurement Scales Health Status Questionnaire 2, for self-assessment of health-related quality of life in all types of osteoporotic patients [25]. It contained 80 questions, later condensed to 59 questions in four domains: physical, psychological, symptoms, and social.

The Quality of Life Questionnaire of the International Osteoporosis Foundation (formerly the European Foundation for Osteoporosis) was developed for patients with vertebral deformities, by a working party of clinicians and quality-of-life specialists from eight countries [5, 18]. A questionnaire of 54 questions (including 6 visual analogue scales) was constructed (Table 3). It was condensed after validation, to 41 questions (Qualeffo-41). Subsequent testing showed that Qualeffo-41 also is suitable for postal administration [26].
Table 3

Qualeffo-41: contents of the quality-of-life questionnaire of the International Osteoporosis Foundation [5, 18]

Domain

Subdomain

Number of questions

Pain

Back pain, sleep disturbance

5

Physical

ADL: dressing, bathing, toilet

4

Jobs around the house: cleaning, cooking, washing dishes, shopping, lifting

5

Mobility: standing up, bending, kneeling, stairs, walking

7

Body image

1

Social

Sport, gardening, hobby, friends

7

General

(Change in) overall QoL

3

Mental

Fatigue, depression, loneliness, energy, cheerfulness, hope, fear

9

The Questionnaire Quality of Life in Osteoporosis (QUALIOST) has a different place, as it was developed as a disease-specific module, in addition to the generic SF-36 questionnaire [27]. It was developed after discussion with patients and contains 23 questions after validation. The patients have to complete 59 questions, but this includes the generic questionnaire SF-36.

Utility

Utility is the value attached to a specific health state. The questions do not address specific complaints but try to value situations. Utility takes the preferences of individuals into account [28]. This is most clear when using the “standard gamble” or “time-trade-off” methods (Fig. 2). The questionnaires or methods are simple and generic: one measure can be used for all diseases. Utility assessment results in one single value for health status ranging from 0 (death) to 1 (perfect health). Utility values can be used to calculate loss or gain of quality-adjusted life years (QALYs). When a certain disease causes a utility change from 1.0 to 0.6 for 1 year, the utility loss is 0.4 and the QALY loss is 0.4. When the same disease causes a utility loss of 0.4 for 6 months or 2 years, the QALY loss is 0.2 or 0.8, respectively. The QALY loss is equivalent to utility loss multiplied by time. Similarly, a gain in QALYs is equivalent to the utility gain multiplied by time.
Fig. 2

Standard gamble (a) and time-trade-off (b) techniques for estimating utility after fracture (adapted from Tosteson ANA, Spine 22:58S–62S, 1997)

Instruments for the calculation of utility and QALYs include the EQ-5D (EuroQol) [15, 29] and the Health Utility Instrument (HUI) [30] questionnaires. The EQ-5D consists of 5 questions concerning 5 domains—i.e., mobility, self-care, usual activities, pain/discomfort, and anxiety/depression—each question having 3 answers. The results are coded with 5-digit numbers, each digit ranging from 1 to 3—e.g., 21121 or 32123. Altogether, 243 codes or health states are possible. Valuations have been done by an expert panel and by the time-trade-off method to obtain utility values [15, 29, 31]. In addition, expert panels and healthy control subjects have been used to obtain utility values for different diseases including osteoporosis. The utility values obtained by experts or healthy control subjects may differ from data obtained in patients.

Two methods, the standard gamble and time-trade-off techniques, are used to directly derive utility values [28, 32]. The standard gamble technique requires an interviewer or computer program. The patient is asked whether he/she prefers certainty (i.e., life with a disability) or a gamble (i.e., a treatment with a chance of perfect health and a risk of death). When the patient accepts an 80% chance of perfect health after treatment versus a 20% risk of death following treatment, the utility is 0.80. When the patient is so ill that he or she would accept the reverse, the utility is 0.20. The interviewer has to go from a 90% to 10% chance of perfect health or death, and balance this with disability risks from 10% to 90%. At the point of indifference—i.e., the patient has no preference between disability and risk-filled treatment—the utility can be calculated. The other technique, time-trade-off, asks for a preference either to live with disability for 10 years or to live with perfect health for X years (less than 10 years). At indifference, the utility can be calculated. For example, if the patient is indifferent regarding the choice between living in perfect health for 3 years or living with a disability for 10 years, the utility is 0.30. Both methods are shown in Fig. 2a, b. An advantage of these two techniques is their simplicity. However, the methods have not been well validated.

Assessment of quality of life in patients with osteoporosis

Quality-of-life data in patients with vertebral fractures

Prevalent vertebral fractures

Data on quality of life in patients with vertebral fractures have been obtained with the six osteoporosis-specific questionnaires and with generic instruments. Most studies were done in patients with existing, i.e., prevalent vertebral fractures. Only some studies concern the quality-of-life change after a new or incident vertebral fracture. All osteoporosis-specific questionnaires discriminated well between patients with prevalent vertebral fractures and control subjects of similar age and sex. Such data usually were obtained in validation studies. Only some questionnaires have been used in large clinical trials involving patients with osteoporosis (defined by BMD T-score <−2.5) either without or with prevalent vertebral fractures.

The Multiple Outcomes of Raloxifene Evaluation (MORE) trial incorporated the Qualeffo-41 and the OPAQ into its protocol. The Qualeffo-41 was used in seven European countries. Quality of life was assessed in 751 postmenopausal women with osteoporosis (i.e., bone mineral density T-score ≤−2.5). At baseline, 449 women had prevalent vertebral fractures (height loss >20%), and 302 women did not [33]. Highly significant differences between the two groups were observed for the domains pain, physical function, social function, and general health, and in the total scores (either the mean of all domain scores or the mean of all individual questions). Mental function was not different between the two groups. A trend toward increasing quality-of-life scores with increasing number of prevalent vertebral fractures was observed for all domains except mental function. An interaction with age was observed indicating higher scores (poorer quality of life) with increasing number of vertebral fractures and increasing age (Fig. 3). The impact of lumbar fractures on quality of life was more than that of thoracic fractures for the domains pain, physical function, and general health, and total score. For example, patients without vertebral fractures had a mean total score of 25.6±14.3 and those with ≥3 thoracic fractures a score of 35.8±19.7, while patients with ≥3 lumbar fractures had a score of 53.2±15.8. The Nottingham Health Profile and EQ-5D were also incorporated in this study, and showed similar but weaker associations.
Fig. 3

Impact of age and number of vertebral fractures on Qualeffo-41 total score. Data from the MORE study (reprinted from Oleksik A et al., J Bone Miner Res 15:1384–1392, 2000)

The OPAQ (version 2; OPAQ2) was used in the MORE study in Australia, Canada, New Zealand, and the United States. The OPAQ2 scores were significantly lower (worse quality of life) in patients with one or more vertebral fractures than in those without, for the domains physical function, emotional status, symptoms, and overall quality of life (one question), whereas social interaction was similar in both groups [34]. Quality of life decreased significantly with increasing number of prevalent vertebral fractures for all domains except social interaction. Lumbar and thoracic fractures had a similar impact on quality of life except for physical function being lower in patients with lumbar fractures (borderline, p=0.055). These data are in agreement with data on days of bedrest and limited activity and on health care use in patients with prevalent vertebral fractures [35, 36]. These parameters show a strong increase with number or severity of vertebral fractures. Data on the relationship between quality of life and vertebral fracture severity (degree of height loss) are not yet available.

Incident vertebral fractures

Reports on the change of quality of life following incident (new) vertebral fractures are scarce. Quality of life was assessed in 157 women with incident vertebral fractures in the MORE study using the OPAQ questionnaire. The patients with incident vertebral fractures had significant quality-of-life loss in the domains physical activity, emotional status, symptoms, and the overall quality-of-life question [34]. In the European part of the MORE trial, the effect of incident vertebral fractures was studied in the group of women with prevalent vertebral fractures at baseline. In this group, 290 women finished 3 years of follow-up without incident vertebral fractures, and 67 had an incident fracture in a previously normal vertebra. Total scores of Qualeffo-41 were significantly higher in the group with incident vertebral fractures than in the women without [37]. This was also true for the physical function domain. About one in three incident vertebral fractures is a clinical fracture, i.e., it leads to immediate clinical attention as the patient visits a general practitioner or hospital and the fracture is diagnosed on a spinal radiograph. Of the 67 patients with incident vertebral fracture, 20 were clinically recognized and 47 were subclinical—i.e., later diagnosed on a scheduled radiograph in the trial. The quality-of-life loss was significant for the subclinical fractures for the total scores and physical function, whereas the clinical fractures caused quality-of-life loss in all domains except mental function. This confirms the impression that subclinical vertebral fractures, even when not immediately diagnosed, cause a decrease in quality of life. The quality-of-life change in persons with an incident vertebral fracture was also assessed within the EPOS study [38]. There were significant differences in total Qualeffo-41 score, SF-12, and EQ-5D between persons with incident vertebral fractures and controls without prevalent vertebral fractures. This difference was mainly caused by patients with an incident vertebral fracture who also had a prevalent vertebral fracture. However, patients with only an incident vertebral fracture were not different from controls.

A direct effect of a therapeutic intervention on quality of life in patients with osteoporosis and vertebral fractures has not been demonstrated. However, when an intervention decreases the incidence of vertebral fractures, this results in a relatively better quality of life, because incident vertebral fractures cause a deterioration of quality of life.

Quality of life following distal forearm fracture

Distal forearm fractures (distal radius fracture, Colles fracture, or wrist fracture) cause immediate pain and loss of function and almost always come to clinical attention. Fracture healing and regain of function are usually good. Quality-of-life data have been obtained in patients with distal forearm fracture using the EQ-5D (EuroQol). A group of 40 patients was followed for 1 year, and the EQ-5D was administered every 3 months [39]. The EQ-5D is a utility instrument, and the quality-of-life loss could thus be expressed in QALYs. Considerable loss was observed in the first 3 months, but recovery was fast. The total quality-of-life loss was 0.05 for 1 year, i.e., 0.05 QALY.

Quality of life following hip fracture

Almost all patients with hip fracture are admitted to hospital, are operated on, and have to be rehabilitated. The quality-of-life loss is considerable. However, the collection of data on quality of life is hampered by the age of the patients. Many patients have impaired cognition before the fracture. In addition, the fracture, hospital admission, and operation may cause disorientation. A self-administered questionnaire may not be feasible since visual impairment often coexists. Interviewer-based questionnaires with simple questions may be more suitable. Morbidity and mortality in patients with hip fracture are considerable. About 20–25% of the patients die within the 1st year following hip fracture. About 30% will not regain mobility [40]. Only half of the patients will walk again and often not to the same degree as before the hip fracture. The recovery and predictors of recovery have been well documented, the best predictor being the prefracture health state [41].

Several studies have documented the loss of physical function. The Yale Health and Aging Project included a prospective study of 120 patients with hip fracture [42]. These were studied before the hip fracture, as part of a 3-year in-home interview, and 6 weeks and 6 months after the hip fracture. A considerable decline occurred in independent dressing, independent transfer, and independent walking across a room. Six months after fracture, only 8% were able to climb stairs compared with 63% before the fracture, and 6% could walk half a mile compared with 41% before the fracture. Long-term follow-up of patients with trochanteric fractures showed a decrease in walking ability and the activities of daily living. After a mean follow-up of 7 years, 36% were able to walk alone outdoors, and 53% were able to dress without help, but 74% could not visit friends [43]. The poor functional outcome was also shown in another study in which hip fracture patients were 4.2 times more likely to be community immobile—i.e., not able to function in the wider community—2 years after the fracture [44]. In addition, hip fracture patients were 2.6 times more likely to be functionally dependent than control subjects.

Similarly, quality of life decreases after hip fracture. In one study, 32 patients with hip fracture and 29 matched controls completed the SF-36 and the OPAQ2 questionnaires at 1 week and 12 to 15 weeks after fracture [45]. The patients with hip fracture had lower baseline scores and a significant decrease in the SF-36 domains physical function, vitality, and social function, and in the OPAQ2 domains physical function, social activity, and general health, while the control group had no change in quality-of-life scores. The decrease in quality of life as measured by the SF-36 was also observed in another study [46]. In this study, patients with hip fracture scored significantly lower on all domains of the SF-36 as compared to controls. In addition, hip fracture patients scored significantly worse on the timed up-and-go test, the Berg Balance scale, and the Frenchay Activities of Daily Living test. Also, the EQ-5D index scores decreased after hip fracture [47]. The scores decreased from 0.78 before the fracture to 0.45 at 1 week, 0.59 at 4 months, and 0.51 at 17 months in survivors of a femoral neck fracture treated with internal fixation. At each follow-up assessment, the mean quality of life was higher for those with healing fractures than for those with fracture healing complications. Patients with comorbidities, patients with pain ratings of 30 mm or less on the Visual Analogue Scale (score 0–100 mm), patients with good mobility, patients with high-level independence in Activities of Daily Living, and patients with an independent living status had better EQ-5D index scores. When comparing 14 survivors of undisplaced (Garden I and II) and 28 survivors of displaced (Garden III and IV) femoral neck fractures, it appears that the rate of fracture-healing complications and reoperations was higher in patients with displaced fractures [48]. In addition, patients with undisplaced fractures scored significantly better on the timed up-and-go test, activities of daily living, need for help from relatives or community services, and quality of life, as measured by the EuroQol. Quality of life after a healed undisplaced fracture was not significantly different before and after fracture, whereas quality of life after a healed displaced fracture was significantly lower at 4, 12, and 24 months after the fracture as compared to before the fracture.

Poor functional outcome was also confirmed in a follow-up study of 102 consecutively admitted patients with hip fracture [49]. Of those who survived 4 months after fracture, 43% reached the prefracture level of walking ability but only 17% reached the same ADL level as before the fracture. The Nottingham Health Profile (NHP) could be obtained in 75% of the patients directly. All dimensions improved between 1 week and 4 months after fracture, but physical mobility, social isolation, sleep, and emotional reactions were still significantly less than in the reference population. A study in 100 hip fracture patients admitted from and discharged to their own homes showed that patients not managing outdoor walking 6 months after the fracture and patients with complications had higher scores on the NHP [50]. The impact of the hip fracture was most obvious in the sections physical mobility and pain of the NHP, as compared with a reference population. In another study it was found that postfracture health perception, as measured by one question of the SF-12, was influenced by income, number of comorbidities, prefracture health perception, postfracture social support, and postfracture function [51].

The effect on quality of life of various fracture types was compared using the SF-36 within the Canadian Multicenter Osteoporosis Study [52]. The time interval between fracture and the study varied between 0 years and more than 10 years. The physical functioning domain was most influenced by hip and pelvis fractures in women, while the role-physical domain was most influenced by hip fractures in men. In general, hip fractures have a severe impact on quality of life, similar to or greater than that of moderate to severe vertebral fractures.

Utility changes after fracture

Utility data for patients with hip and/or vertebral fracture are presented in Table 4. Most data have been obtained with the time-trade-off method. These indicate that utility is lower in patients after hip fracture and patients after vertebral fracture. Utility was worse in patients with a disabling hip fracture, a hip and vertebral fracture combined, or multiple vertebral fractures. Other methods (standard gamble, health utility index) yielded different results. Two studies in older women who were asked to value different outcomes for hip or vertebral fractures showed that older women without osteoporotic fractures judged a worse quality of life after fracture than the patients themselves [32, 53].
Table 4

Utility values for patients with hip fracture and/or vertebral fracture(s) derived from patients with hip fracture and/or vertebral fracture, aged control subjects or expert panels

Author

Subject

Fracture

Method

Utilitya

Tosteson [30]

Patients

1–2 year hip fracture

Time-trade-off

0.48 (0.32–0.64)

1–2 year hip and vertebral fracture

Time-trade-off

0.32 (0.10–0.53)

1–2 year vertebral fracture

Time-trade-off

0.80 (0.68–0.91)

Cranney [10]

Patients

Recent hip fracture

Feeling thermometer

0.71 (0.50–0.85)

Standard HUI 2, gamble

0.91 (0.75–1.0)

0.67 (0.53–0.89)

Cranney [10]

Patients

Recent vertebral fracture

Feeling thermometer

0.76 (0.50–0.95)

Standard gamble

0.84 (0.50–1.0)

HUI 2

0.79 (0.25–0.92)

Gabriel [32]

Patients

Hip fracture

Time-trade-off

0.65±0.45

Disabling hip fracture

Time-trade-off

0.28

Multiple vertebral fracture

Time-trade-off

0.68±0.40

Gabriel [32]

Older women

Hip fracture

Time-trade-off

0.28±0.37

Multiple vertebral fracture

Time-trade-off

0.31±0.38

Salkeld [53]

Older women

Fear of falling

Time-trade-off

0.67

“Good” hip fracture

Time-trade-off

0.31

“Bad” hip fracture

Time-trade-off

0.05

NOF expert panel [54]

Expert panel

Hip fracture, 1st year

Expert estimate

0.53

Dutch consensus panel [55]

Expert panel

Hip fracture, 1st year

Expert estimate

0.61

aMean (95% confidence interval), or mean ± SD, or mean

The utility loss after fracture has also been estimated by expert panels. An expert panel of the National Osteoporosis Foundation estimated the utility (QALY) loss after hip fracture in the 1st year at 0.4681, mainly due to hospital and nursing home stay, decreasing in the 2nd year to 0.1695 QALY. The QALY loss after distal radius fracture was 0.0464 in the 1st year and 0.0006 in the 2nd year [54]. However, dependency after this fracture type could cause a QALY loss of 0.30 in the 2nd year. The average QALY loss after a vertebral fracture was estimated at 0.0502 during the 1st year and 0.0490 QALY in the 2nd year. However, a fracture with severe pain could cause a QALY loss of up to 0.5000. The average QALY loss may appear low, but only one of three such deformations comes directly to clinical attention. A substantial decrease of physical function occurs after one severe deformation or three light to moderate deformations [36]. A Dutch expert panel reached similar estimates of QALY loss to those of the NOF panel after distal radius fracture and vertebral fracture, but their estimate for loss of quality of life after hip fracture was lower than that of the NOF panel [55].

Implications of quality-of-life studies in patients with osteoporosis

Quality-of-life data obtained in patients with osteoporotic fractures show that loss of quality of life is more severe in patients after hip or multiple vertebral fractures than in patients with a single vertebral fracture or distal radius fracture. The longer after the fracture, the more quality of life improves, but quality of life is not completely restored. This indicates that prevention of osteoporotic fractures is of utmost importance in order to maintain quality of life. Concerning the burden of disease, utility studies show that the burden of osteoporosis can be compared to that of other chronic diseases.

The value of quality-of-life instruments in clinical practice in individual patients is uncertain. It might remind the doctor and the patient of the many possible complaints. However, the variation coefficient probably will be too big to detect changes during follow-up. Although some cultural adaptation usually is done during translation of a questionnaire, multicenter studies suggest differences in domain scores between countries—e.g., between southern and northern Europe—but the data are insufficient up till now.

With regard to clinical trials, it is recommended that one include both a disease-specific and a generic instrument. Using a generic instrument, the burden of disease can be compared and utility may be assessed. The latter is important for cost-effectiveness estimates.

Conclusion

In recent years, quality-of-life studies have taught us about the complaints and preferences of patients with osteoporotic fractures. Suitable methods are available to assess quality of life in patients with osteoporotic fractures. Quality of life can be measured to compare the effect of different treatments in patients with osteoporosis. The assessment of quality of life is also useful in epidemiologic studies to estimate the burden of disease and to evaluate the cost-effectiveness of different treatment scenarios in health care policy. Further studies are required to assess the influence of comorbidity in patients with vertebral and hip fractures and to evaluate cultural differences in quality-of-life scores. Simple instruments should be developed for the assessment of quality of life in elderly patients with hip fractures.

Aknowledgements

Natasja van Schoor has been funded by an unrestricted grant of Wyeth Research, Collegeville, Pennsylvania, USA.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2004