Maturitas 53 (2006) 245–251

PIXI bone density screening for osteoporosis in postmenopausal women Ross Lawrenson a,∗ , Peter Nicholls b , Rebecca Rivers-Latham a , Tessa Brown c , Jonathan Barnardo d , Richard Gray e a

Postgraduate Medical School, University of Surrey, Guildford, Surrey, UK b University of Aberdeen, Aberdeen, UK c Woodbridge Hill Surgery, Guildford, Surrey, UK d St. Lukes Surgery, Guildford, Surrey, UK e Royal Surrey County Hospital, Surrey, UK

Received 5 October 2004; received in revised form 3 May 2005; accepted 17 May 2005

Abstract Objectives: The aim of this study was to evaluate a pragmatic screening programme for osteoporosis based on the identification of known risk factors. A secondary aim was to assess the validity of peripheral instantaneous X-ray imager (PIXI) scanning against dual energy X-ray absorptiometry (DEXA) in women identified as having osteopenia. Methods: A cross-sectional two stage screening programme. The study was carried out in 14 practices in Surrey. Women aged 60–80 years of age were screened with a questionnaire. Those identified with one or more risk factors were offered a PIXI scan of the ankle in their own surgery. Those with an intermediate score on PIXI scan were offered a DEXA scan of hip, spine and forearm. Results: Four thousand six hundred and forty-six women completed questionnaires, 2688 had a PIXI scan and 553 were found to be at high risk of osteoporosis. Multivariate analysis identified the three most important risk factors associated with increased risk of osteoporotic fracture as age, a previous fracture and the presence of a stooped posture. Hormone replacement therapy (HRT) was shown to be protective. Twenty three percent of women with an intermediate score on PIXI scan were found to have osteoporosis on DEXA scan of hip and spine. Conclusions: PIXI scanning proved acceptable, practicable but only had moderate comparability with DEXA. The findings suggest that patients over the age of 60 years with a history of a fracture or evidence of spinal collapse are likely to have osteoporosis and should be offered screening. HRT past the menopause would seem to confer benefit and the recent reduction in its use may lead to increasing numbers of women suffering osteoporotic fractures. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Osteoporosis; Densitometry; Hormone replacement therapy; Family practice; Mass screening

∗

Corresponding author. Tel.: +44 1483 579927; fax: +44 1483 300359. E-mail address: [email protected] (R. Lawrenson).

0378-5122/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.maturitas.2005.05.004

246

R. Lawrenson et al. / Maturitas 53 (2006) 245–251

1. Background In the UK, osteoporosis causes 200,000 fractures each year, causing severe pain and disability to individual patients at an estimated annual cost to the National Health Service of over £940 million. More than one-third of adult women will sustain one or more osteoporotic fractures in their lifetime [1]. The National Osteoporosis Society estimated that 120 hip fractures per primary care organisation of 100,000 patients will cost £21,000 per fracture at an average cost of £2,580,000 per primary care organisation [2]. Hip fractures are also associated with a significantly increased mortality of between 20% and 35% within the first year [3]. There are good studies showing the association between low bone mass density (BMD) and subsequent fractures [4–8]. The identification of older women at risk of a fractured neck of femur has been seen by many as increasingly important. It has been shown that there are effective interventions available, which will reduce the osteoporotic fracture rate of older women, including the use of hormone replacement therapy (HRT) [9] and bisphosphonates [10]. The key challenge to the medical community is how to most easily identify those women at high risk who would benefit from treatment. Objective measurement to diagnose osteoporosis in high-risk women can be carried out using dual energy X-ray absorptiometry (DEXA) scanning. DEXA is relatively expensive with a UK cost of around £70 per scan. The National Osteoporosis Society states that a primary care trust serving a population of 100,000 would require approximately 1000 DEXA scans per year [2]. Mass screening by DEXA scanning is not recommended without some selection of the target population. A cheaper alternative is a peripheral instantaneous X-ray imager (PIXI) scan, which is portable and can be used in the community. We have shown it is easily possible for a nurse with 2–3 days training to screen 40 patients a day which equates to approx £3–20 p of nurse time per scan. Given that a PIXI scanner costs between £10,000 and 15,000, if used regularly the cost of each scan will be of the order of £5 per patient. PIXI produces an objective measure of peripheral bone density and because of its ease of use and relatively low cost a number of UK National Health Service Trusts have set up a diagnostic service using PIXI. However, the correlation between PIXI and DEXA needs to be

considered before using PIXI as a screening tool in primary care. The use of an algorithm based on known risk factors in conjunction with PIXI scans could limit the requirement for DEXA. The aim of this study was to test the utility of such an algorithm in UK general practice. A secondary aim was to test the validity of PIXI scans in diagnosing osteoporosis against a gold standard of DEXA.

2. Methodology The study was carried out by the Guildford and Waverley and East Waverley Primary Care Groups (now the Guildford and Waverley PCT). Fourteen of the 29 practices took part in the study. A protocol was agreed between primary and secondary care, which incorporated a selective case finding strategy. This involved targeted screening and self-assessment through patients filling in a questionnaire sent to them by post. It was decided to use a questionnaire as it was believed much of the data thought to be relevant could not be reliably obtained from the patient’s computer records. All women aged 60–80 years of age were identified using the practice database and sent a questionnaire, which listed some key risk factors and demographic details (Table 1). In association with the questionnaire patients were asked to sign a consent form to allow their data to be used in the study. Women were excluded from being sent a question for three reasons: (1) they were terminally ill, (2) they were already diagnosed and/or being treated for osteoporosis and (3) they were in a nursing home or in residential care. These patients were felt to be at risk [11] and therefore did not require prior investigation. Recruitment of patients took place during the month of September 2001 with PIXI screening being undertaken over a 14-month period starting in October 2001 and finishing in January 2003. The entire research project was undertaken using the computerised patient care system within the primary care group. The results of the questionnaires were entered on to a Microsoft Access database and women with one or more risk factors for osteoporotic fracture were invited for a PIXI scan. Questions numbers 2 and 11 (Table 1) were for information only and not considered to be risk factors when deciding which patients should be invited

R. Lawrenson et al. / Maturitas 53 (2006) 245–251

247

Table 1 The questionnaire sent to all women aged 60–80 years of age at risk of osteoporotic fracture The questionnaire 1. Did you have your menopause before the age of 45? 2. Are you currently on hormone replacement therapy (HRT)? 3. Did your period stop at any other time for longer than 6 months (pregnancy excluded)? 4. Have you recently completed a course of steroids? For example, Prednisolone 5. Have you fallen in the last 6 months? 6. Have you fractured/broken your hip, wrist or vertebrae since the age of 45? 7. Did you mother fracture/break their hip and suffer from osteoporosis (brittle bones)? 8. Would you say that you walk with a stoop (curvature of the spine) or a loss of height (more than one inch)?

(21) (11) (3.3) (3.4) (17) (12.3) (12.3) (17.9)

9. Do you suffer from any of the following? (a) Long-standing liver disease (b) Long-standing kidney disease (c) Coeliac or any other disease of malabsorption (d) Thyroid or parathyroid

(0.3) (0.4) (1.9) (9.9)

10. Do you consider yourself housebound, e.g. unable to go shopping, walk the dog or visit relations? 11. Have you ever smoked?

(3.4) (42)

Four thousand six hundred and forty six replied. Number in ( ) is the percent of women who answered “yes” to each question.

for screening. PIXI scanning of the ankle was carried out in the GP surgery by an osteoporosis project nurse who was trained in this method of screening. PIXI scanning measures peripheral BMD and is recorded as the patient’s T-score. The results were sent to their GP using the computerised shared patient care system. Women were divided into three categories: Low risk Moderate risk High risk

Better than −0.6 T-score −0.6 to −1.6 T-score Worse than −1.6 T-score

According to the manufacturer, these categories are equivalent to T-scores of >−1, −1 to −2.5 and 75

1409 1739 1669 2590

958 1164 1091 1433

584 699 684 721

61 (10.4%) 98 (14.0%) 153 (22.4%) 241 (33.4%)

116 (19.9%) 217 (31.0%) 215 (31.4%) 196 (7.4%)

407 (69.7%) 384 (54.9%) 316 (46.2%) 284 (39.4%)

Total

7407

4646

2688

553 (21.0%)

744 (27.7%)

1391 (51.7%)

questionnaire and identifies a high proportion (21%) of women in the study who had their menopause before 45 years of age. Data also indicated that 17% of screened patients over 60 years of age had fallen in the last 6 months and 18% said they had lost height or walked with a stoop. Very few patients reported having chronic kidney or liver disease, a malabsorption problem or were using steroids. 11% of women aged >60 years were taking HRT at the time of the screening. There was no difference between the uses of HRT in those with one or more risk factors as compared to those with no risk factors. Table 2 gives details of the study response stratified by age group. 51.7% were identified as having a T-score >−0.6 and were classified as low risk. The prevalence of osteopenia/osteoporosis is shown to increase with age with 33.4% of screened women over the age of 75 being identified as at high risk (T < −1.6). Table 3 shows the results of the multivariate analysis. The factors significantly associated with an increased risk are a history of a fracture OR = 1.6 (95% CI 1.30–2.02), walking with a stoop OR = 1.5 (95% CI 1.23–1.84), liver disease OR = 3.11 (95% CI 1.05–9.21) and increasing age. Women aged 75–80 years were identified as being three times more likely to be at increased risk of osteoporosis than those aged 60–64 years. Women taking HRT had a significantly reduced risk of osteoporosis with an odds ratio (OR) 0.42 (95% CI 0.27–0.64). Only screening those with a history of a prior fracture or evidence of kyphosis or loss of height would reduce the number of women to be screened by half. Seven hundred and forty four/2688 patients who had a PIXI scan had a score of −0.6 to −1.6. Of these, 420 had a DEXA scan result recorded from all three sites

(hip, forearm and spine). Only patients with an intermediate classification on PIXI were invited for a DEXA. Each of the 11 possible PIXI scores has been compared with the median DEXA score for hip, spine and forearm with the attendant 95% CI. The results for hip and spine are reproduced in Figs. 1 and 2. The positive correlation between PIXI and DEXA were statistically significant for all five categories although this was relatively weak with R2 less than 2 in all cases (Table 4). When considering all 420 PIXI compared with hip DEXA the R2 was 0.1180 and p < 0.005. The percentage of patients with DEXA < −2.5 is also presented in Table 4 and shows that based on the “gold standard” of DEXA of the hip and spine the false negative rate of PIXI is at least 23%. Table 3 A multivariate analysis of risk factors associated with high-risk Tscores on PIXI screening Variables

Odds ratio

CI95

Age (60–64) Age (65–69) Age (70–74) Age (>75) Early menopause Periods stopped for more than 6 months Recent course of steroids Fallen in last 6 months Fracture since age of 45 years Mother suffered from osteoporosis Walk with a stoop Liver disease Coeliac disease Thyroid disease Considered themselves housebound Former or current smoked Currently taking HRT

1.0 1.25 1.86 2.90 1.19 0.84 0.98 1.08 1.62 0.95 1.50 3.11 1.56 1.12 1.01 0.88 0.42

Reference 0.91–1.73 1.36–2.54 2.14–3.93 0.98–1.46 0.54–1.30 0.66–1.45 0.88–1.33 1.30–2.02 0.74–1.22 1.23–1.84 1.05–9.21 0.96–2.54 0.87–1.44 0.69–1.47 0.73–1.06 0.27–0.64

R. Lawrenson et al. / Maturitas 53 (2006) 245–251

249

4. Discussion

Fig. 1. The median DEXA score for total hip compared with PIXI scores from −0.6 to −1.6 with attendant 95% CI (number 420). The expected DEXA for each PIXI score is also illustrated.

Fig. 2. The median DEXA score for spine compared with PIXI scores from −0.6 to −1.6 with attendant 95% CI (number 420). The expected DEXA for each PIXI score is also illustrated.

Table 4 Percentage of 420 women with an intermediate score on PIXI (−0.6 to −1.6) of the ankle who are found to have a false negative measurement (T < −2.5) on DEXA at different sites

DEXA of hip (total score) DEXA of spine DEXA of forearm Worst of three Worst of hip and spine

False negative rate (%)

R2

p-Value

8.1 21.0 29.3 40.2 23.1

0.118 0.133 0.162 0.167 0.163

0.016 0.006 0.001 0.001 0.001

Also, presented are the correlation coefficients for PIXI vs. DEXA for each of the five categories with relevant p-values.

This study demonstrated that a screening programme for osteoporosis can be carried out by a Primary Care Organisation and provided in a range of practices. The cut off point for referral to a medical practitioner for consideration for treatment was based on existing guidelines. Using the screening questionnaire led to approximately 2200 women per 100,000 practice population being screened with PIXI. This is a slightly greater than that calculated by a smaller study from Keele University [13]. The study showed that the three most important risk factors associated with increased risk of osteoporotic fracture are age, a previous fracture and the presence of a stooped posture. Early menopause, halted periods, recent use of steroids, a fall in the last 6 months, mother having a fracture, coeliac or thyroid, kidney disease, housebound and smoking were not associated with a low BMD. A single stage population based screening programme in general practice in The Netherlands identified similar findings to those in this study with age, previous fracture, kyphosis and a low BMI being the key risk factors [14]. Other factors, which were not identified in this study as significant but have been in other studies include steroids, family history of fracture, low body weight and smoking [5,6,13]. Only using the two variables of a previous fracture and kyphosis would reduce the number of women requiring BMD screening to 1000 investigations per 100,000 practice population. In this study, the recent use of steroids was not identified as being associated with a low BMD on PIXI screening. In the Royal College of Physicians recent paper on glucocorticoids, they were identified as a risk factor for fracture at the hip and spine. It was also stated that the risk of fracture is higher in steroid induced osteoporosis than in postmenopausal osteoporosis [15]. The identification of steroids as not being significant in this study could be chance due to the low numbers of patients identified in the questionnaire as having recently taken a course of steroids. It may also be due to the way the question was asked, with no record of the length of the steroids’ use or dose being recorded. Patients who recorded that they had liver disease were three times more likely to have a low BMD. Only 15 patients responded “yes” to this question of

250

R. Lawrenson et al. / Maturitas 53 (2006) 245–251

which 7 had a low BMD. It is not known whether the liver disease was associated with alcohol abuse in these patients, but they are a group of patients who may warrant further study [16]. HRT was shown to be protective against a high risk of an osteoporosis with an OR of 0.42. This suggests that prolonged use of HRT, past the age of menopause is worthwhile. We had 59 high-risk women over the age of 70 years taking HRT. There has been a reduction in the use of HRT following studies reporting women taking HRT to have a higher risk of developing coronary heart disease and breast cancer [17,18]. Re-examination of these studies has brought the findings into doubt but if there is a sustained reduction in the long-term use of HRT this will result in an increased number of women with osteoporosis, increasing their risk of osteoporotic fracture by 65% [19]. The strengths of this study are that it was a population based screening investigation, which included a large number of patients with complete data obtained by using a questionnaire. One concern was that 37% of women sent a questionnaire did not respond. This is similar to a similar study in general practice in The Netherlands [20]. We do not now whether these women who did not complete a questionnaire were different to those who did respond, but it is likely a proportion were at high risk of osteoporotic fracture. Limitations include a concern about the generalisability of the study to other parts of the UK as the fracture risk in the Surrey population may not be representative of the UK as a whole. Surrey has the least deprivation of almost any county in the UK and another UK primary care study has shown that less deprived women have a significantly higher BMD and are less likely to have osteoporosis [21]. However, another study from Nottingham using PIXI screening, classified 47.9% of women aged over 60 years as normal, 27.7% as osteopenia and 21.0% as osteoporotic which are very similar to our findings of 51.7%, 27.7% and 21.0%, respectively. They found 9.3% were current users of HRT compared to 11% in Surrey [22]. This suggests that our results are comparable with other studies in the UK. The lack of data in our study on BMI was unfortunate and would have been helpful in clarifying whether the findings on BMI from The Netherlands are also applicable here. Another concern would be the reliability and validity of PIXI scanning to assess risk of osteoporotic

fracture. GE Medical, the supplier’s of the peripheral instantaneous X-ray imager detail the accuracy of the PIXI scanner as being −1.6 on PIXI and it would be worthwhile assessing this before setting up a local service. Also, the variation of both the PIXI and DEXA scores were considerable as shown in Figs. 1 and 2 and indicated by the R2 and this should be borne in mind by those considering PIXI as a screening tool. However, it should also be remembered that a low score on peripheral scanning is associated with four times the risk of a subsequent fracture [6] and justifies treating this high-risk group.

5. Conclusion Based on these findings, we suggest that if undertaking a screening programme for osteoporosis in general practice that the best indicators of risk are age, past history of fracture or evidence of stoop. A further indicator that should be considered is a low BMI but other risk factors suggested by the College of Physicians Guidelines are of little use in primary care. Rather than using a postal questionnaire it may be better for primary care doctors to identify patients from their computerised records and use these as a basis for invitation for PIXI screening. A combination of PIXI scan and a history of symptoms is suggested as being a cost effective way of identifying women in whom treatment should be considered. However, there is a strong possibility of false negative results from PIXI. We suggest that Primary Care Organisations should set up a specialist osteoporosis assessment clinic where high-risk women could be referred for advice and screening. This could include PIXI scanning although continued access to DEXA would also be needed.

R. Lawrenson et al. / Maturitas 53 (2006) 245–251

Acknowledgements We would like to thank all the practices that took part in our study, Graham Simmons, Data Manager of Guildford and Waverley Primary Care Trust for all his computer support through the project and Alex Burn, Health Promotion Manager for Guildford and Waverley Primary Care Trust for her support of the programme. We also would like to thank Proctor and Gamble, Eli Lilly and Shire Pharmaceuticals Ltd. who supported our study.

References [1] Royal College of Physicians. Osteoporosis: clinical guidelines for prevention and treatment; 1999. Updated 2000. [2] The National Osteoporosis Society. Primary care strategy for osteoporosis and falls: a framework for health improvement programmes implementing the national service framework for older people; October 2002. [3] Goldacre MJ, Roberts SE, Yeates D. Mortality after admission to hospital with fractured neck of femur: database study. BMJ 2002;325(7369):868–9. [4] Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: report of a WHO study group. World Health Organisation Tech Rep Ser 994;843:1–129. [5] Chesnut 3rd CH. Osteoporosis, an underdiagnosed disease. JAMA 2001;286(22):2865–6. [6] Siris ES, Miller PD, Barrett-Connor E, et al. Identification and fracture outcomes of undiagnosed low bone mineral density in postmenopausal women: results from the National Osteoporosis Risk Assessment. JAMA 2001;286(22):2815–22. [7] Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996;312(7041):1254–9. [8] Cummings SR, Black DM, Nevitt MC, et al. Bone density at various sites for prediction of hip fractures. The Study of Osteoporotic Fractures Research Group. Lancet 1993;341(8837):72–5. [9] Levine JP. Long-term estrogen and hormone replacement therapy for the prevention and treatment of osteoporosis. Curr Women’s Health Rep 2003;3(3):181–6. [10] Cranney A, Waldegger L, Zytaruk N, et al. Risedronate for the prevention and treatment of postmenopausal osteoporosis. In: Cochrane database of systematic reviews, vol. 4; 2003. CD004523.

251

[11] Ooms ME, Vlasman P, Lips P, Nauta J, Bouter LM, Valkenburg HA. The incidence of hip fractures in independent and institutionalized elderly people. Osteoporos Int 1994;4:6– 10. [12] StataCorp. Stat statistical software: release 7.0. College Station, Texas: Stata Corporation; 2001. [13] Thomas E, Richardson JC, Irvine A, Hassell AB, Hay EM. Osteoporosis: what are the implications of DEXA scanning ‘high risk’ women in primary care? Family Pract 2003;20(3):289–93. [14] Versluis RG, Papapoulos SE, de Bock GH, Zwinderman AH, Petri H, van de Ven CM. Clinical risk factors as predictors of postmenopausal osteoporosis in general practice. Br J Gen Pract 2001;51(471):806–10. [15] Royal College of Physicians. Glucocorticoid induced osteoporosis: guidelines for prevention and treatment; 2003. [16] Collier JD, Ninkovic M, Compston JE. Guidelines on the management of osteoporosis associated with chronic liver disease. GUT 2002;50(Suppl. I):i1–9. [17] Writing Group for the Women’s Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women’s Health Initiative randomized controlled trial. JAMA 2002;288(3):321–3. [18] Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women Heart and Estrogen/progestin Replacement Study (HERS) Research Group. JAMA 1998;280(7):605–13. [19] Yates J, Barrett-Connor E, Barlas S, Chen YT, Miller PD, Siris ES. Rapid loss of hip fracture protection after estrogen cessation: evidence from the National Osteoporosis Risk Assessment. J Family Pract 2004;53(6):444–7. [20] Versluis RG, Petri H, van de Ven CM, Scholtes AB, Papapoulos SE, Springer MP. Prevalence of osteoporosis in postmenopausal women in family practice. Ned Tijdschr Geneeskd 1999;143(1):20–4. [21] Pearson D, Taylor R, Masud T. The relationship between social deprivation, osteoporosis, and falls. Osteoporos Int 2004;15(2):132–8. [22] Masud T, Jordan D, Hosking DJ. Distal forearm fracture history in an older community-dwelling population: the Nottingham Community Osteoporosis (NOCOS) study. Age Ageing 2001;30:255–8. [23] GE Medical Systems LUNAR. PIXI: peripheral DEXA bone densitometer; 2001. [24] Pearson D, Masud T, Sahota O, Earnshaw S, Hosking D. A comparison of calcaneal dual-energy X-ray absorptiometry and calcaneal ultrasound for predicting the diagnosis of osteoporosis from hip and spine bone densitometry. J Clin Densitom 2003;6(4):345–52.