Skip Navigation


Age and Ageing Advance Access originally published online on June 9, 2008
Age and Ageing 2008 37(5):536-541; doi:10.1093/ageing/afn128
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
37/5/536    most recent
afn128v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Chen, J. S.
Right arrow Articles by Sambrook, P. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, J. S.
Right arrow Articles by Sambrook, P. N.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Copyright © The Author 2008. Published by Oxford University Press on behalf of the British Geriatrics Society.

Fracture risk assessment in frail older people using clinical risk factors

Jian Sheng Chen1, Judy M. Simpson2, Lyn M. March1, Ian D. Cameron3, Robert G. Cumming4, Stephen R. Lord5, Markus J. Seibel6 and Philip N. Sambrook1

1 Institute of Bone and Joint Research, University of Sydney, St Leonards, NSW 2065, Australia
2 Department of Public Health and Community Medicine, University of Sydney, Sydney, NSW 2006, Australia
3 Rehabilitation Studies Unit, University of Sydney, Ryde, NSW 2112, Australia
4 Centre for Education and Research on Ageing, University of Sydney, Concord, NSW 2139, Australia
5 Prince of Wales Medical Research Institute, UNSW, Randwick, NSW 2031, Australia
6 Anzac Research Institute, University of Sydney, Concord Hospital, Concord, NSW 2139, Australia

Address correspondence to: Jian Sheng Chen. Tel: (+61) 2 9926 7328; Fax: (+61) 2 9906 1859. Email: jschen{at}med.usyd.edu.au


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 Key points
 Conflicts of interest...
 Supplementary data
 References
 
Background: this study aims to develop and evaluate a simple fracture risk index for use in frail older people.

Methods: clinical risk factors were assessed at baseline for 2,005 older people (473 males, 1,532 females; mean age 85.7 years, SD 7.1 years) living in aged-care facilities. Fractures were ascertained for 2 years from baseline. Cox regression model was used to identify significant risk factors for fracture. Hazard ratios (HRs) from the model were assigned as weights. The risk index was calculated by multiplying the weights of all risk factors.

Results: during a mean follow-up of 1.64 years, 401 fractures occurred in 338 participants. Significant independent clinical risk factors for fracture were institution type, balance, history of previous fracture, cognitive function, number of medications, weight and lower leg length (n = 1,813). The index was capable of identifying higher-risk individuals, with almost an 8-fold increase in the risk of fracture for residents from the lowest 15% to the highest 18% of the score. Among 1-year survivors, a high score (≥15) indicated approximately a one-in-six chance of fracture, while a low score (<8) indicated only a one-in-forty chance of fracture within a year. The area under the receiver operating characteristic (ROC) curve was 0.69 (95% CI: 0.65–0.72) and 0.68 (95% CI: 0.65–0.71) for identifying someone who would have a fracture in 1 and 2 years respectively.

Conclusions: this risk index could identify individuals at higher fracture risk among institutionalised older people, and thus, could help to rationalise the provision of fracture prevention programs in this population.

Keywords: fracture, aged, risk assessment, risk factors, elderly


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 Key points
 Conflicts of interest...
 Supplementary data
 References
 
Osteoporotic fractures occur in about 0.5% of the population per annum in countries such as the United States, European countries and Australia [1]. Most of the fractures occur in older age and result in a significant increase in morbidity and impairment of quality of life in this elderly population. Residents of aged-care facilities contribute to a significant proportion of the problem. In the Auckland Hip Fracture Study, 42% of hip fracture patients aged 60 years or older are residents in these facilities [2].

Most fracture risk assessments are focussed on older people living in the community [3–9]. Bone fragility, as measured by bone mineral density (BMD) or quantitative ultrasound (QUS), has been shown to predict fractures among the community dwelling elderly [4, 10, 11]. Factors such as older age, female gender, fracture after 50 years of age, maternal hip fracture, neuromuscular, cognitive and visual impairments, low weight and current smoking also contribute to the risk of fracture, independent of BMD or broadband ultrasound attenuation (BUA) [6, 12]. Combining measures of bone fragility with clinical risk factors has been shown to give better prediction of future fracture risk [8, 13]. However, a fracture risk index developed for community dwelling subjects may not be useful for the frail elderly living in residential care facilities, for whom risk factors may differ.

Institutionalised older people have a higher fracture incidence than community dwellers [14, 15], but risk factors for this population have been less well studied. Older people living in residential care have increased levels of chronic illness, medication use, and cognitive, vision, strength and balance impairments. The importance of bone fragility as a fracture risk factor is likely to be reduced, while risk of falling and the impact of a fall become relatively more important in this population [8, 16, 17]. In frail older people, measuring BMD is impractical, and it has been suggested that using BUA measurements in assessing risk of fracture is also unlikely to have a large benefit [3]. Markers of bone turnover also do not seem to predict risk of fracture among the frail elderly who are at high risk of falls [18]. Few fracture risk assessments have been developed for these older people based on clinical risk factors [15]. Moreover, in frail older people, multiple risk factors are likely to be present and inter-related. A risk assessment without weighting for the relative importance of each risk factor will not reflect the true likelihood of a fracture and may not fully differentiate those high-risk individuals.

The aim of this study was to develop and evaluate a simple fracture risk index based only on clinical risk factors, which takes into account the relative importance of each risk factor. This was undertaken in a prospective cohort study of older people living in nursing homes and intermediate care facilities in Australia.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 Key points
 Conflicts of interest...
 Supplementary data
 References
 
Subjects
This analysis was conducted using all 2,005 participants in the Fracture Risk Epidemiology in the Frail Elderly (FREE) Study [19–21].

Baseline and outcomes assessments
The following clinical risk factors were assessed: demographics (age, sex, ethnicity), type of residence (nursing home and intermediate care facilities), use of walking aids, history of smoking, cognitive function [the Standardised Mini-Mental State Examination (SMMSE)], co-morbidities (a modification of the Implicit Illness Severity Scale), use of medications, osteoarthritis of the knee, Parkinson's disease, stroke, total hip replacement, total knee replacement, incontinence, history of previous fracture (since age 50 years), falls in previous year, postural stability, reaction time, visual contrast sensitivity, sit-to-stand ability, weight and lower leg length (an index of mature skeletal stature). Details of these measurements are given in previous publications [19–21].

Fracture data were collected for 2 years from baseline and validated by x-ray reports as previously reported [20]. The diagnosis of clinical vertebral fractures required subjects to have radiological evidence in association with acute back pain at a corresponding level (but a previous normal X-ray at that vertebral level was not required).

Statistical analysis
A Cox regression model was developed to identify risk factors for first fracture. Participants who died (n = 354 by 1 year, 689 by 2 years) or were lost to follow-up (n = 17) were treated as censored. Risk factors in the final model were selected based on their significance level (P < 0.05) and ease of assessment in a primary-care setting. If there was a significant interaction between two variables, a single new variable was created to reflect the fact that the effect of one variable was modified by the other variable. In order to develop a simpler model, categories with similar risk effects in a particular variable were combined.

The exponential of the coefficient of each risk factor from the selected model, i.e. the hazard ratio (HR), was assigned as a weight if a resident had that factor. A weight of 1 was given if a resident did not have that risk factor. The fracture risk index was calculated by multiplying the weights of all risk factors (as shown in Results). Therefore, the fracture risk index is a relative risk to those without the risk factors. This methodology is based on the Cox proportional hazards regression equation: h(t) = h(0)exp(β1x1 + ... + βkxk) = h(0)expβ1x1*... *expβkxk where h(t) is a hazard rate and h(0) is a baseline hazard function for an individual with x1 = ... = xk = 0.

The index was tested by comparing ratios of the scores with incidence-rate ratios (IRR) of the observed fracture rates to see how well the index predicted risk of fractures. Probabilities of fractures in the first year for different scores were calculated, as well as sensitivities, specificities and likelihood ratios of detecting a fracture in a year. Finally, areas under the receiver operating characteristic (ROC) curves were calculated using logistic regression models.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 Key points
 Conflicts of interest...
 Supplementary data
 References
 
The study cohort comprised 1,107 intermediate care facility residents and 898 nursing home residents. Most participants were very old and frail with a mean age of 85.7 years (SD 7.07 years), with 70% using walking aids. Only 15% of residents were taking vitamin D and/or calcium supplements. Baseline characteristics of study subjects are presented in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1. Characteristics of study population

 
During a mean follow-up of 1.64 years, 401 low trauma fracture events were documented in 338 participants, giving an overall fracture rate of 12.2 per 100 person-years (9.4 and 14.1 per 100 person-years for residents of nursing homes and intermediate care facilities, respectively). Of these 338 subjects, 84.9% had one fracture event, 12.5% had two and 3.6% had three. The fracture sites were: 132 hip, 82 vertebral, 49 pelvic, 35 wrist, 32 humeral, 27 rib, 17 femoral shaft and 60 miscellaneous fractures.

Only 1,813 of the 2,005 participants (1,064 in intermediate care facilities and 749 in nursing homes) had data recorded for all risk factors identified in the final multivariate regression model. Table 2 shows these significant independent predictors and their assigned weights (i.e. HR). The fracture risk index for an individual was calculated by multiplying the weights for each risk factor. For example, the index would be 12.28 (4.52 x 1 x 1 x 2.14 x 1.27) for an intermediate care resident with good balance, SMMSE >23, number of medications ≤6, weight <53 kg and lower leg length ≥50.5 cm. The fracture risk index was therefore an incidence-risk ratio compared to a resident without the risk factors.


View this table:
[in this window]
[in a new window]

  		Table 2. Hazard ratios (HR) from a multivariate Cox regression model and weighting for fracture risk index

 
The 1,813 participants were divided into six roughly equal-sized groups according to their scores (rounding the cut-off points to the nearest integer). Table 3 presents ratios of the scores and IRRs of the observed fracture rates for the different index groups. Values of all IRRs were similar to the corresponding score ratios, indicating that the risk index corresponded well to the observed fracture rate. For example, the fracture rate was 2.9 per 100 person-years for participants with a mean score of 3.06, compared to a rate of 23.1 per 100 person-years for those with a mean score of 26.40, a multiplicative increase of 7.9. The 29% of participants who had a score <8 accounted for only 13% of all fractures, while 57% of fractures occurred in the 35% of participants with a score ≥15 (calculated from Table 3). These results show that the selected model works well for this particular dataset.


View this table:
[in this window]
[in a new window]

  		Table 3. The fracture risk index score ratios and incidence-rate ratios (IRR) by fracture risk index

 
Figure 1 (see Appendix 1 in the supplementary data on the Journal's website http://www.ageing.oxfordjournals.org) presents fracture rates by the index groups for all participants, and those who survived >1 year or >2 years. Fracture rates for residents of nursing homes and intermediate care facilities are presented in Figure 2 by quintile of risk score (Appendix 1 in the supplementary data on the Journal's website http://www.ageing.oxfordjournals.org). These figures show that the index clearly differentiates fracture risk among the residents, regardless of how long they survive or where they live.

Table 4 (see Appendix 1 in the supplementary data on the Journal's website http://www.ageing.oxfordjournals.org) shows the number of participants who did and did not suffer a fracture in the first year among the 1,493 who survived at least 1 year, as well as probability of a fracture, and gives sensitivities, specificities and positive likelihood ratios according to the five cut-off points. Among 1-year survivors the 1-year fracture risk was only 2.5% for a resident with a score <8, but was 15.9% for those with a score ≥15.

The area under the ROC curve of the risk assessment was 0.69 (95% CI: 0.65–0.72) and 0.68 (95% CI: 0.64–0.72) for identifying someone—among all participants and among those who survived for more than a year, respectively—who would have a fracture in 1 year. For identifying someone who would have a fracture in 2 years, the area under the ROC curve was 0.68 (95% CI: 0.65–0.71) among all participants and 0.67 (95% CI: 0.63–0.71) among those who survived for more than 2 years.

Due to concern about possible bias from including only those vertebral fractures that came to clinical attention, we repeated the multivariate analysis in Table 2 using only the first non-vertebral fracture (n = 286) and the same risk factors. The results were very similar for most risk factors, while HRs for weight, balance and institution type were somewhat larger; the discrimination capability was also very similar.


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
Key points
 Conflicts of interest...
 Supplementary data
 References
 
We have developed a fracture risk index based on a large prospective cohort of residents from randomly selected aged care facilities in a defined geographical area. The index is derived from seven easily assessed risk factors and weighted according to the relative importance of each risk factor. The index was capable of identifying higher-risk individuals from this high-risk group over a 2-year period, with almost an 8-fold increase in the risk of fracture for residents from the lowest 15% to the highest 18% of the score.

Age was one of the most important components of a fracture risk index in other studies [7, 8, 13, 15, 22], but was not an independent risk factor for the FREE participants. The effect of age on risk of fracture is likely to be through its correlation with other risk factors. Alternatively, the lack of effect of age may be because admission to aged care facilities is determined by the level of disability, not age, so that only very disabled 70-year-olds, for example, are admitted. QUS measure was not included as part of the fracture risk index as we wanted it to be derived only from simple clinical assessments. Similarly, visual contrast sensitivity was not included because it is not easily assessed, particularly in residents with cognitive impairment. We found that the effect of postural stability (i.e. balance) on risk of fracture was modified by type of residence, which was factored into our risk index. The different effects of postural stability on risk of fracture may be partly explained by the different levels of supervision provided in the two types of facility. Residents who were extremely frail (i.e. nursing home residents with a ‘not capable’ balance grade) were at low risk of fracture. This might be because these residents ambulated only with assistance, were under heavy supervision and may have sustained falls with lesser impact (e.g. fall from a chair) because they were incapable of standing unaided.

It is not possible to compare our index directly with the scoring algorithm developed by Girman et al. [15] since neither of the studies measured the same risk factors. In that study of 1,427 white female nursing home residents, the scoring algorithm developed by classification and regression-tree methodology contained age, weight, height, mobility, dementia, falls in the last 6 months, urinary incontinence and activities of daily living score. Our fracture risk index had a slightly better fracture discriminating ability than the algorithm used by Girman et al. (i.e. the area under ROC curve = 0.69 ± 0.018 and 0.68 ± 0.015 for identifying someone who would have a fracture in 1 and 2 years respectively in the FREE study versus 0.63 ± 0.043 for identifying someone who would have a fracture in 1.5 years in the Girman study). A strength of our index is its potential for greater generalisability, given that it was derived from a sample that includes both males and females, and residents of both intermediate care facilities and nursing homes, and that it has a similar discriminating ability at 1 and 2 years. It is the only published fracture risk index whose score for residents with all seven risk factors is a relative risk for fracture compared to residents with none of those risk factors. It can therefore be used to compare the fracture risk for residents with different risk factors.

We expect that this index will be applicable to other institutionalised elderly populations. However, further evaluation is needed as the performance of the index has not yet been tested. This may be of particular interest in other populations outside Australia, and where calcium and vitamin D supplementation is more commonly given [23]. The index should not be used to exclude residents who may be at low risk of fracture from an effective universal intervention program such as calcium and vitamin D supplements [23]. To maintain the utility of the index, this study has selected only those risk factors that can be accurately measured in other settings. The reliability of these tests, which could increase transportability of the index, has been established in previous studies [24–26]. In this study, misclassification bias is possible since many vertebral fractures do not come to clinical attention. However, secondary analysis excluding vertebral fractures showed that any bias in HRs for risk factors was towards the null value of 1. Moreover, caution should be applied for an individual risk assessment since the index is derived from aggregated estimates.

In conclusion, the fracture risk index proposed here could identify individuals at higher risk of fracture among institutionalised older people, and thus, could help to rationalise the provision of fracture prevention programs in this population.


   Key points
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
Conflicts of interest...
 Supplementary data
 References
 

  • Clinical risk factors for fracture included institution type, balance, history of fracture, cognitive function, number of medications, weight and lower leg length.
  • Fracture risk index derived from these risk factors was capable of identifying higher-risk individuals for fracture among older people living in both intermediate care facilities and nursing homes.
  • The index score provides a relative risk for fracture.


   Conflicts of interest declaration
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 Conflicts of interest...
Supplementary data
 References
 
The authors have nothing to disclose.


   Supplementary data
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 Conflicts of interest...
 Supplementary data
References
 
Supplementary data for this article are available online at http://ageing.oxfordjournals.org.


   Acknowledgements
 
Funding for this work was provided by the Australian National Health and Medical Research Council, and Osteoporosis, Australia. The authors would like to thank Ms J Schwarz (Research Co-ordinator) and Ms J Makaroff (Research Assistant) for their efforts in co-ordinating the study and collecting the data, and all the participants for giving their time in providing valuable information. We gratefully acknowledge the support we received from the staff members in the participating institutions.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 Conflicts of interest...
 Supplementary data
 References
 

  1. Access Economics Pty Limited. The Burden of Brittle Bones: Costing Osteoporosis in Australia. In: Prepared for Osteoporosis Australia (2001).
  2. Butler M, Norton R, Lee-Joe T, et al. The risks of hip fracture in older people from private homes and institutions. Age Ageing (1996) 25:381–5.[Abstract/Free Full Text]
  3. McGrother CW, Donaldson MM, Clayton D, et al. Evaluation of a hip fracture risk score for assessing elderly women: the Melton Osteoporotic Fracture (MOF) study. Osteoporos Int (2002) 13:89–96.[CrossRef][Web of Science][Medline]
  4. Khaw KT, Reeve J, Luben R, et al. Prediction of total and hip fracture risk in men and women by quantitative ultrasound of the calcaneus: EPIC-Norfolk prospective population study. Lancet (2004) 363:197–202.[CrossRef][Web of Science][Medline]
  5. De Laet CE, Van Hout BA, Burger H, et al. Hip fracture prediction in elderly men and women: validation in the Rotterdam study. J Bone Miner Res (1998) 13:1587–93.[CrossRef][Web of Science][Medline]
  6. Cummings SR, Nevitt MC, Browner WS, Study of Osteoporotic Fractures Research Group. Risk factors for hip fracture in white women. N Engl J Med (1995) 332:767–73.[Abstract/Free Full Text]
  7. Colon-Emeric CS, Pieper CF, Artz MB. Can historical and functional risk factors be used to predict fractures in community-dwelling older adults? development and validation of a clinical tool. Osteoporos Int (2002) 13:955–61.[CrossRef][Web of Science][Medline]
  8. Black DM, Steinbuch M, Palermo L, et al. An assessment tool for predicting fracture risk in postmenopausal women. Osteoporos Int (2001) 12:519–28.[CrossRef][Web of Science][Medline]
  9. Dargent-Molina P, Piault S, Breart G. A comparison of different screening strategies to identify elderly women at high risk of hip fracture: results from the EPIDOS prospective study. Osteoporos Int (2003) 14:969–77.[CrossRef][Web of Science][Medline]
  10. Bauer DC, Gluer CC, Cauley JA, et al, A prospective study. Study of Osteoporotic Fractures Research Group. Broadband ultrasound attenuation predicts fractures strongly and independently of densitometry in older women. Arch Intern Med (1997) 157:629–34.[Abstract/Free Full Text]
  11. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ (1996) 312:1254–9.[Abstract/Free Full Text]
  12. Dargent-Molina P, Favier F, Grandjean H, et al. Fall-related factors and risk of hip fracture: the EPIDOS prospective study. Lancet (1996) 348:145–9.[CrossRef][Web of Science][Medline]
  13. Burger H, De Laet CE, Weel AE, et al. Added value of bone mineral density in hip fracture risk scores. Bone (1999) 25:369–74.[CrossRef][Web of Science][Medline]
  14. Cali CM, Kiel DP. An epidemiologic study of fall-related fractures among institutionalized older people. J Am Geriatr Soc (1995) 43:1336–40.[Web of Science][Medline]
  15. Girman CJ, Chandler JM, Zimmerman SI, et al. Prediction of fracture in nursing home residents. J Am Geriatr Soc (2002) 50:1341–7.[CrossRef][Web of Science][Medline]
  16. Hayes WC, Myers ER, Morris JN, et al. Impact near the hip dominates fracture risk in elderly nursing home residents who fall. Calcif Tissue Int (1993) 52:192–8.[CrossRef][Web of Science][Medline]
  17. Visentin P, Ciravegna R, Uscello L, et al. Site-specific relative risk of fractures in the institutionalized elderly. Gerontology (1995) 41:273–9.[Web of Science][Medline]
  18. Chen JS, Seibel MJ, Zochling J, et al. Calcaneal ultrasound but not bone turnover predicts fractures in vitamin D deficient frail elderly at high risk of falls. Calcif Tissue Int (2006) 79:37–42.[CrossRef][Web of Science][Medline]
  19. Lord SR, March LM, Cameron ID, et al. Differing risk factors for falls in nursing home and intermediate-care residents who can and cannot stand unaided. J Am Geriatr Soc (2003) 51:1645–50.[CrossRef][Web of Science][Medline]
  20. Sambrook PN, Cameron ID, Chen JS, et al. Influence of fall related factors and bone strength on fracture risk in the frail elderly. Osteoporos Int (2007) 18:603–10.[CrossRef][Web of Science][Medline]
  21. Zochling J, Sitoh YY, Lau TC, et al. Quantitative ultrasound of the calcaneus and falls risk in the institutionalized elderly: sex differences and relationship to vitamin D status. Osteoporos Int (2002) 13:882–7.[CrossRef][Web of Science][Medline]
  22. Dargent-Molina P, Douchin MN, Cormier C, et al. Use of clinical risk factors in elderly women with low bone mineral density to identify women at higher risk of hip fracture: The EPIDOS prospective study. Osteoporos Int (2002) 13:593–9.[CrossRef][Web of Science][Medline]
  23. Chapuy MC, Arlot ME, Duboeuf F, et al. Vitamin D3 and calcium to prevent hip fractures in the elderly women. N Engl J Med (1992) 327:1637–42.[Abstract]
  24. Han TS, Lean ME. Lower leg length as an index of stature in adults. Int J Obes Relat Metab Disord (1996) 20:21–7.[Web of Science][Medline]
  25. Lord SR, Clark RD, Webster IW. Physiological factors associated with falls in an elderly population. J Am Geriatr Soc (1991) 39:1194–200.[Web of Science][Medline]
  26. Molloy DW, Alemayehu E, Roberts R. Reliability of a standardized mini-mental state examination compared with the traditional mini-mental state examination. Am J Psychiatry (1991) 148:102–5.[Abstract/Free Full Text]
Received 8 August 2007; accepted in revised form 26 February 2008.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
37/5/536    most recent
afn128v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Chen, J. S.
Right arrow Articles by Sambrook, P. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, J. S.
Right arrow Articles by Sambrook, P. N.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?