WITH the aging population in many Western countries,

Transcription

1 Journal of Gerontology: MEDICAL SCIENCES 2006, Vol. 61A, No. 10, Copyright 2006 by The Gerontological Society of America Falls Risk and Functional Decline in Older Fallers Discharged Directly From Emergency Departments Melissa A. Russell, 1,2 Keith D. Hill, 2 Irene Blackberry, 2 Lesley L. Day, 3 and Shyamali C. Dharmage 1 1 Centre of MEGA Epidemiology, School of Population Health, The University of Melbourne, Parkville, Victoria, Australia. 2 Preventative and Public Health Division, National Ageing Research Institute, Parkville, Victoria, Australia. 3 Monash University Accident Research Centre, Monash University, Clayton, Victoria, Australia. Background. There is currently no standard approach to falls risk assessment and management for older fallers presenting to the emergency department (ED) who are discharged directly home. Hence, this study was conducted to describe the prevalence of falls risk factors associated with older fallers presenting to the ED and to identify the factors associated with postdischarge decline in function in this group. Methods. This cross-sectional study was performed with 300 community-dwelling individuals, aged 60 years or older, admitted to the ED following a fall, and discharged directly home. A home-based assessment after ED discharge was performed, which included the prevalence of falls risk factors, identification of functional decline, and objective measurements of balance, gait, depression, and falls efficacy. Results. Fall-related injuries were sustained by 91% (95% confidence interval [CI], 87.2% 94.0%) of participants presenting to the ED. The most common falls risk factors identified in the home assessment were polypharmacy (79.0%, 95% CI, 73.9% 83.5%), home hazards (76.0%, 95% CI, 70.8% 80.7%), decreased balance (61.3%, 95% CI, 55.6% 66.9%), and arthritis (61.3%, 95% CI, 55.6% 66.9%). A decline in function was reported by 35% of participants (95% CI, 29.6% 40.7%). Sustaining a fracture, functional independence before the fall, being female, depression, and slower Timed Up and Go (TUG) scores were associated with a decline in function ( p,.05). Conclusion. Older fallers discharged directly from the ED have a high prevalence of falls risk factors and are at risk of functional decline. WITH the aging population in many Western countries, the number of older people sustaining falls is escalating. By 2020 an estimated four million older Americans will fall each year (1), and approximately 12% of these falls will result in serious injury (2). Many older fallers sustaining an injury will present to a hospital emergency department (ED), with falls accounting for 62% of all nonfatal injury-related ED visits in persons aged 65 years and older (3). Community-dwelling older fallers presenting to the ED are at high risk of further falls, with 52% sustaining falls in the subsequent 12 months (4). Eighty-two percent of older fallers presenting to the ED after a fall are treated and discharged directly home from the ED (3). Despite this, there is currently no standard approach to falls risk assessment and management for older fallers presenting to the ED and discharged home. Presentation to the ED is a valuable opportunity to identify persons at increased risk and to develop management plans to reduce the risk of further falls. However, patient assessment time is limited in EDs, and decisions need to be made quickly. Knowledge regarding the falls risk profile and functional limitations encountered by this population is required to guide post ED discharge management programs in such high-pressure environments. Although summary information is available in one study (5), there has been no published comprehensive analysis of falls risk factors for community dwellers attending an ED and discharged home after a fall. In addition to the risk of further falls, older people presenting to the ED have been found to be at risk of a decline in short-term functional ability; approximately one fifth of older people presenting to the ED with a minor injury experienced an ongoing decline in function 3 months after discharge (6). The factors associated with functional decline among fallers presenting to the ED, and their predictive value, are relatively unknown, thereby limiting the opportunity to improve outcomes. Such research is particularly pertinent, as previous studies have shown that a decline in function can continue for 12 months in persons sustaining fractures (7), and an ongoing decline in physical activity after injury is associated with the onset of depression (8). The objectives of this study were to: (a) describe the characteristics and prevalence of falls risk factors among community-dwelling older fallers who have presented to the ED and are discharged directly home, and (b) identify the factors associated with a short-term decline in ability to perform Activities of Daily Living (ADLs) after discharge home. METHODOLOGY Study Design and Sample This is a cross sectional study of baseline data of a randomized controlled trial assessing the effectiveness of 1090

2 FALLS RISK AND FUNCTION IN ED OLDER FALLERS 1091 a falls prevention intervention. Data were collected between January 2003 and December Participants were recruited through hospital staff from the EDs of seven acute hospitals in Melbourne, Australia. The inclusion criteria were presenting to the ED with a fall, aged 60 years or older, discharged directly home following emergency care, living in the community, and being able to walk independently. A standard falls definition was used (9). Potential participants were excluded if they were cognitively impaired, defined as having an Abbreviated Mental Test Score less than 7 (10). Written informed consent was gained from all participants, and the project was approved by the Ethics Committees at the participating hospitals. Data Collection Methods A post ED assessment was conducted using a structured protocol by a physiotherapist, occupational therapist, or doctor (allocation depending on time availability) at the participant s own home. The median number of days between the ED presentation and the assessment was 20 (interquartile range: 12 33). A review of the participant s medical records was undertaken to check for the diagnosis of concurrent acute medical conditions (infection, electrolyte imbalance, dehydration, or a prefall orthopedic injury) underlying the fall that caused the ED presentation. Location of the Fall and Injury Sustained A self-report of location and injuries sustained in the fall resulting in the ED presentation was obtained from each participant, and the most serious injury sustained is reported here. Falls Risk Factors The presence or absence of risk factors was based on participant self-report, researcher observation, or objective measure. The risk factors selected for this study were based on two published reviews of falls risk factors in community-dwelling adults (11,12). The inter-rater reliability of the risk factor assessments was tested on 20 consecutive participants, with 1 week between two standardized assessments, one conducted by the physiotherapist and the other by the doctor. The average Kappa statistic for the falls risk factor variables was 0.60 (95% confidence interval [CI], ). Self-reported risk factors included: Age Having a history of falls (defined as sustaining at least one other fall in the previous 12 months, in addition to the one that brought the person to the ED) Arthritis of lower limbs or spine Incontinence Dizziness at any time in the past 3 months Somatosensory impairment (defined as constant loss of sensation in the lower limbs) Visual deficit (defined as trouble seeing objects clearly [visual acuity], trouble judging distance [depth perception], or a history of visual impairment but no vision testing in past 2 years) Use of a walking aid indoors or outdoors Depression (defined by a score of 6 or above on the Geriatric Depression Scale Short-Form [GDS-SF]) (13) Prefall functional impairment (defined by needing assistance with any personal or domestic ADLs [including showering, dressing, cooking, housework, shopping]) prior to the fall. The assessor observed or measured the presence of the following risk factors: Polypharmacy (.3 medications) all recorded from bottles and/or packets Psychotropic medications (sedatives, antidepressants, or antipsychotics) Cardiac medications (Class 1A antiarrythmics, digoxin, diuretics) Balance deficit was measured using the Step Test (14) and Functional Reach Test (FR) (15). Balance impairment was defined as having a worst Step Test leg score of 10 or fewer steps per 15 seconds and an FR below 28 cm. These cutoffs are two standard deviations below the mean scores for a healthy group of older women (16). Home hazards (defined by having loose mats, cords across walkways, clutter, or rails required on stairs or in the bathroom) Gait impairment was assessed using the Timed Up and Go (TUG) test (17), and gait velocity over a 5 meter walkway. Gait impairment was defined as having a TUG above seconds and usual gait speed less than meter/min. The cutoff for the TUG has been recommended previously (18), and the cutoff for gait velocity is two standard deviations below the mean score for a nonimpaired group of older people (19). Falls Efficacy The Modified Falls Efficacy Scale (MFES) (20) asks participants to rate their current level of confidence in performing 14 functional activities without falling. Higher scores on the 0 10 scale are indicative of less fear of falling. Outcome Measure A decline in function was defined as requiring assistance with at least one domestic and/or personal ADL that the participant did not require assistance with prior to the fall. Analysis The percentage of participants with each falls risk factor in the post ED discharge assessment and 95% CI were calculated. For non-normally distributed continuous measures, medians and interquartile ranges are presented. The association between the factors measured during the assessment and decline in function postfall were examined using multiple logistic regression analysis. Odds ratios and 95% CIs are presented. The variables included in this analysis were gender, polypharmacy, depressive symptoms (21), history of previous falls (22), fracture (7,23), bruising (6), sprain and/or strain, laceration, dislocation, concussion, low falls efficacy (24), increased age, prefall independence (25), balance measures, gait measures, presence of arthritis, diagnosis of a new underlying medical condition in the

3 1092 RUSSELL ET AL. ED, and living alone. All continuous measures were divided into quintiles and evaluated for a linear association between the measure and functional decline. The number of days between the ED discharge and assessment and the occurrence of falls between the ED discharge and assessment were considered to be potential confounders. Univariate analysis was performed between all independent variables and a decline in function. The associations were re-examined while adjusting for the two confounders. The independent variables found to be statistically significant (p,.1) in the univariate analysis were then entered into a multiple logistic regression model. Likelihood ratios were then used to construct the strongest multiple logistic regression model. STATA (Survey Design and Analysis Services Pty Ltd, Blackburn, Victoria, Australia) was used for all analyses. RESULTS Study Sample During the study period, 480 eligible patients expressed an initial interest in the study to ED department staff; 300 consented and subsequently completed the baseline assessment, 153 did not wish to participate, and 27 could not be contacted after discharge. All participants were directly discharged home from the ED; however, 5 (1.7%) participants were admitted to the hospital at a later date for management of their fall injury. The mean age of the 300 study participants was 76.9 years (95% CI, years). Two hundred nine (69.7%) participants were female, and 140 (46.7%) lived alone. In the days between the ED presentation and assessment, 30 (10.0%, 95% CI, 6.8% 14.0%) participants fell again. Of the 106 participants seen within 2 weeks of the ED presentation, only 4 (3.8%, 95% CI, 1.0% 9.4%) fell. The rate of falls in the period from discharge to assessment was 7.3 falls per 1000 persondays (95% CI, ). Fall Characteristics, Injuries, and Risk Factors Two hundred seventy-three participants (91%; 95% CI, 87.2% 94.0%) reported sustaining an injury in the fall resulting in the ED presentation. The most common injury reported was fracture, and over a third of participants reported that the fall occurred inside their own home (Table 1). The mean falls risk factor per participant was 7.3 (95% CI, 7.0% 7.6%) from the 17 risk factors assessed. The most common risk factors for falls, in descending order, were polypharmacy, home hazards, balance deficit, and arthritis (Table 1). Fifteen (5.7%) of the 261 participants for whom medical records were available had an acute medical diagnosis (infection 3.4%, prefall orthopedic injury 1.5%, and dehydration 0.8%) recorded as contributing to their fall. Functional Decline One hundred five participants (35%, 95% CI, 29.6% 40.7%) reported an ongoing decline in function at the time of the assessment compared to prefall levels. Univariate analysis revealed being female, sustaining a fracture, having a higher GDS-SF score, and having a slower TUG score increased the odds of a decline in function postfall. Functional impairment prior to the fall, faster gait speed, a higher Step Test score, a higher FR, and a higher MFES lowered the odds of a decline in function postfall (Table 2). Adjustment for the number of days between ED presentation and assessment, and the occurrence of falls between the ED presentation and assessment, did not statistically alter the significance of the odds ratios in the univariate analysis (results not shown). Additionally, neither of these two variables was independently associated with a decline in function postfall. In the multiple logistic regression model, the variables continuing to be associated with a decline in function were sustaining a fracture, being female, having a slower TUG score, and having a higher GDS score. Negatively associated was functional impairment prefall (Table 2). Gait speed, FR score, Step Test score, and MFES were no longer associated after adjustment for the other variables (p..05). The odds ratios for the variables included in the final model are fully adjusted for the number of days between ED presentation and assessment, and the occurrence of falls between the ED presentation and assessment. DISCUSSION Fall Characteristics, Injuries, and Risk Factors This study provides a unique detailed analysis of the falls risk profile of community-dwelling older fallers presenting to the ED and discharged directly home. A large number of participants sustained an injury (91%) in the fall resulting in the ED presentation, with fracture the most common injury reported (32.7%). Almost two thirds of participants sustained at least one additional fall in the 12 months prior, and the number of falls risk factors per person was high. Both of these factors have been reported previously as associated with risk of further falls (11,26). The rate of falls in the average of 20 days following ED discharge was high (7.3/ 1000 person-days), relative to those of community-dwelling older people (1.9/1000 person-days) and institutionalized older people (4.4/1000 person-days) (27). Differences between study definitions of falls risk factors make comparison of the prevalence of falls risk factors between different populations difficult. However, comparison can be made between the objective measures taken in this study and those of other defined populations of older people. Figure 1 shows a comparison between the objective measures taken in this study, healthy older people (16,19,20), and older people attending a falls clinic [a high falls risk population (28)]. The summary statistics of the objective measures taken with this study s participants are similar to those obtained with the high risk group of older people attending a falls clinic and differ notably from that of healthy older people (Figure 1). This study has some important implications for health professionals. First, older fallers discharged directly home from the ED appear to be at high risk of further falls, considering the rate of falls in the post ED discharge period, highlighting a need for risk assessment after presentation to the ED. Preferably such an assessment would take place as

4 FALLS RISK AND FUNCTION IN ED OLDER FALLERS 1093 Table 1. Circumstances, Outcomes of the Fall Resulting in the ED Presentation, and the Falls Risk Factors Found After Discharge From the ED Falls-Related Variables Location of the fall Percentage (n) N ¼ % CI Objective Measure: Mean Indoors 35.7 (107) NA NA Outdoors at home 29.0 (87) NA NA Community 35.3 (106) NA NA Injury Fracture 32.7 (98) NA NA Contusion/abrasion 25.7 (77) NA NA Laceration 14.3 (43) NA NA Sprain/strain 13.0 (39) NA NA None 9.0 (27) NA NA Dislocation 4.0 (12) NA NA Concussion 1.3 (4) NA NA Falls risk factors Polypharmacy 79.0 (237) NA NA Home hazards 76.0 (228) NA NA Balance deficit 61.3 (184) NA NA FR (cm) Step Test (steps/15 s) Arthritis 61.3 (184) NA NA History of falls in past 12 mo 61.0 (183) NA NA Gait impairment 50.7 (152) NA NA Gait velocity (m/min) Stride length (cm) TUG (s) 15.7* * Use of walking aid 48.3 (145) NA NA Prefall functional impairment 44.3 (133) NA NA Age 80 y 35.0 (105) NA NA Psychotropic medications 33.7 (101) NA NA Visual deficit 30.3 (91) NA NA Incontinence 27.0 (81) NA NA Cardiac drugs 26.7 (80) NA NA Dizziness 26.7 (80) NA NA Depression 26.0 (78) NA NA Somatosensory impairment 15.3 (46) NA NA Other related variable MFES 7.9* * Notes: *Median and interquartile range presented. ED ¼ emergency department; CI ¼ confidence interval; FR ¼ Functional Reach; TUG ¼ Timed Up and Go; MFES ¼ Modified Falls Efficacy Scale; NA ¼ not applicable. 95% CI soon as would be practical, as in this study the number of people sustaining further falls increased with any delay in assessment. Second, the falls risk factors of medications, home hazards, and decreased balance were very common. Considering that previous research has identified that addressing these factors can prevent falls (29) a post ED assessment including these factors could readily underpin an evidence-based falls management program. Functional Decline In the post ED assessment, over one third of participants reported an ongoing loss of function after the fall. As would be expected, fracture was strongly associated with shortterm functional decline postfall. Also associated with functional decline were prefall functional independence, being female, depression (higher GDS-SF score), and poorer postfall mobility (slower TUG). The unexpected nature of the association between prefall functional independence and functional decline postfall could be due to a number of factors. The patients who are functionally impaired prefall may be more likely to be admitted, and hence excluded from this study. Alternatively, independent older people may be more likely to attribute their decline in function to the single event (the fall) compared with less functionally independent older people, who may not distinguish any postfall loss in function from an already apparent decline. This finding requires further research. However, this result does highlight the importance of assessing for short-term assistance needs in prefall functionally independent older people, in addition to identifying those already experiencing prefall functional impairment. These findings should be interpreted in the context of the potential limitations of the study. The sample may be a selective group, as some people may have not wished to

5 1094 RUSSELL ET AL. Risk Factor Characteristics Table 2. Regression Analysis Between Post ED Discharge Assessment Factors and a Decline in Function No Decline in Function: Prevalence (%) or Mean 6 SD (N ¼ 195) Decline in Function: Prevalence (%) or Mean 6 SD (N ¼ 105) Univariate ORs (95% CI) Multivariate ORs (95% CI) Female 62.6% 82.9% 2.89 ( )*,y 2.58 ( ) Age z ( ) Injury None 11.3% 4.8% 1.00 Fracture 21.5% 53.3% 5.87 ( )*,y 3.76 ( ) Contusion/abrasion 31.8% 14.3% 1.06 ( ) Laceration 16.9% 9.5% 1.33 ( ) Sprain/strain 12.8% 13.3% 2.46 ( ) Dislocation 4.1% 3.8% 2.20 ( ) Concussion 1.5% 1.0% 1.46 ( ) Falls risk factors GDS-SF z ( )*,y 1.40 ( ) Balance: Step Test (steps/15 s) z ( ) y Prefall functional impairment 51.3% 31.4% 0.44 ( )*,y 0.25 ( ) Balance: FR (cm) z ( ) y Medication use (.3) 77.4% 81.9% 1.31 ( ) Arthritis 59.5% 64.8% 1.25 ( ) Mobility: TUG (s) z ( ) y 1.47 ( ) Mobility: gait velocity (m/min) z ( ) y History of falls in past 12 mo 61.3% 60.1% 0.98 ( ) Other potentially related variables Diagnosis of acute medical condition 6.98% 3.37% 0.47 ( ) MFES z ( ) y Living alone 47.2% 45.7% 0.94 ( ) Days between ED presentation and assessment z ( ) 0.92 ( ) 1 fall since the ED visit 11.8% 6.7% 0.53 ( ) 0.62 ( ) Notes: *p,.05. p,.1. Quintiles used. ED ¼ emergency department; SD ¼ standard deviation; OR ¼ odds ratio; CI ¼ confidence interval; GDS-SF ¼ Geriatric Depression Scale Short-Form; FR ¼ Functional Reach; TUG ¼ Timed Up and Go; MFES ¼ Modified Efficacy Scale. participate or been missed by busy ED staff identifying participants. A second limitation is that we were unable to determine whether decreased mobility and depression preceded or followed the fall and decline in function. Decreased mobility and depression are risk factors for falls (11) but may also result from an injury and lack of recovery (8). Potentially there may be a more complex cause-and-effect relationship between the factors. The fall and subsequent problems may also prompt over- or underestimation of the level of prefall independence and hence, decline in function attributed to the fall. Summary Older fallers being discharged directly home from the ED have a high prevalence of risk factors for further falls, which are potentially modifiable; have usually sustained an injury; and are at risk of functional decline. A post ED discharge assessment, performed as soon as practical, could address the risk of further falls and need for short-term assistance. Figure 1. Comparison between older fallers being discharged directly from the emergency department (ED), those attending a falls clinic, and healthy older nonfallers. *Falls clinic (28); y healthy female (16); z healthy older (19); healthy older (20); k medians and interquartile ranges presented for ED samples. FR ¼ Functional Reach; TUG ¼ Timed Up and Go; MFES¼ Modified Falls Efficacy Scale; CI ¼ confidence interval. ACKNOWLEDGMENTS This work was supported by the Australian Government Department of Veterans Affairs and the Victorian Department of Human Services. Address correspondence to Melissa Russell, Bachelor of Physiotherapy, National Ageing Research Institute, Poplar Road, Parkville, Victoria, Australia, m.russell@nari.unimelb.edu.au

6 FALLS RISK AND FUNCTION IN ED OLDER FALLERS 1095 REFERENCES 1. Englander F, Hodson TJ, Terregossa RA. Economic dimensions of slip and fall injuries. J Forensic Sci. 1996;41: Campbell AJ, Borrie MJ, Spears G, Jackson SL, Brown JS, Fitzgerald JL. Circumstances and consequences of falls experienced by a community population 70 years and over during a prospective study. Age Ageing. 1990;19: Centre for Disease Control and Management. Public health and ageing: nonfatal injuries among older adults treated in hospital emergency departments-united States, MMWR Morb Mortal Wkly Rept. 2003;52: Close J, Ellis M, Hooper R, Gluscksman E, Jackson SH, Swift CG. Prevention of falls in the elderly trial (PROFET): a randomised controlled trial. Lancet. 1999;353: Lightbody E, Watkins C, Leathley M, Sharma A, Lye M. Evaluation of a nurse-led falls prevention programme versus usual care: a randomized controlled trial. Age Ageing. 2002;31: Shapiro MJ, Partridge RA, Jenouri I, Micalone M, Gifford D. Functional decline in independent elders after minor traumatic injury. Acad Emerg Med. 2001;8: Scaf-Klomp W, van Sonderen E, Sanderman R, Ormel J, Kempen GI. Recovery of physical function after limb injuries in independent older people living at home. Age Ageing. 2001;30: Scaf-Klomp W, Sanderman R, Ormel J, Kempen GI. Depression in older people after fall-related injuries: a prospective study. Age Ageing. 2003;32: The prevention of falls in later life. A report of the Kellogg International Working Group on Prevention of Falls by the Elderly. Dan Med Bull. 1987;34(S4): Hodkinson H. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing. 1972;1: Guideline for the prevention of falls in older persons. American Geriatrics Society, British Geriatrics Society, and American Academy of Orthopaedic Surgeons Panel on Falls Prevention. J Am Geriatr Soc. 2001;49: Moreland J, Richardson J, Chan D, et al. Evidence based guidelines for the secondary prevention of falls in older adults. Gerontology. 2003;49: Sheikh JI, Yesavage JA. Geriatric depression scale (GDS). Recent evidence and development of a shorter version. Clin Gerontol. 1986; 5: Hill KD, Bernhardt J, McGann AM, Maltese D, Berkovits DF. A new test of dynamic standing balance for stroke patients: reliability, validity and comparison with healthy elderly. Physiother Can. 1996;48: Duncan PW, Weiner DK, Chandler J, Studenski S. Functional reach: a new clinical measure of balance. J Gerontol Med Sci. 1990;45A: M192 M Isles RC, Choy NL, Steer M, Nitz JC. Normal values of balance tests in women aged J Am Geriatr Soc. 2004;52: Podsiadlo D, Richardson S. The timed Up and Go : a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991; 38: Bischoff HA, Stahelin HB, Monsch AU, et al. Identifying a cut-off point for normal mobility: a comparison of the timed up and go test in community dwelling and institutionalised elderly women. Age Ageing. 2003;32: Morris, M, Matyas TA, Iansek R, Summers JJ. Temporal stability of gait in Parkinson s disease. Phys Ther. 1996;76: Hill KD, Schwarz JA, Kalogeropoulos AJ, Gibson SJ. Fear of falling revisited. Arch Phys Med Rehabil. 1996;77: Stel VS, Smit JH, Pluijm SM, Lips P. Consequences of falling in older men and women and risk factors for health service use and functional decline. Age Ageing. 2004;33: Tinetti ME, Williams CS. The effect of falls and fall injuries on functioning in community-dwelling older persons. J Gerontol Med Sci. 1998;53A:M112 M Kempen GI, Ormel J, Scaf-Klomp W, Van Sonderen E, Ranchor AV, Sanderman R. The role of perceived control in the process of older peoples recovery of physical functions after fall-related injuries: a prospective study. J Gerontol Psychol Sci Social Sci. 2003;58B: Cumming RG, Salkeld G, Thomas M, Szonyi G. Prospective study of the impact of fear of falling on activities of daily living, SF-36 scores, and nursing home admission. J Gerontol Med Sci. 2000;55A: M299 M Grisso JA, Schwarz DF, Wolfson V, Polansky M, LaPann K. The impact of falls in an inner-city elderly African-American population. J Am Geriatr Soc. 1992;40: Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. New Engl J Med. 1988; 319: Rubenstein LZ, Josephson KR. The epidemiology of falls and syncope. Clin Geriatr Med. 2002;18: Hill K, Black K, Vrantsidis F, NiHill M, Smith R, Dorevitch M. Evaluation of falls clinics and development of a measurement and outcomes framework and review of a movement disorders clinic National Ageing Research Institute. Last accessed on 1/10/2005. Available at: falls_clinic_project_report.pdf 29. Gillespie LD, Gillespie WJ, Robertson MC, Lamb SE, Cumming RG, Rowe BH. Interventions for preventing falls in elderly people (Cochrane Review). In: the Cochrane Library, Issue 3, Chichester, UK: John Wiley and Sons, Ltd. Received December 4, 2005 Accepted March 26, 2006 Decision Editor: Luigi Ferrucci, MD, PhD