In the Public Domain

Journal ofGerontology: MEDICAL SCIENCES 1999, Vol. 54A, No.2, M83-M88

Use of an Ambulation Assistive Device Predicts Functional Decline Associated With Hospitalization Jane E. Mahoney," Mark A. Sager,':" and Muhammad Jalaluddirr' 'Geriatric Research, Education and Clinical Center, William S. Middleton Memorial Veterans Hospital, Madison, Wisconsin. Departments of' Medicine, 3Preventive Medicine, and 4Biostatistics, The University of Wisconsin-Madison.

Background. Loss of functional independence occurs frequently with hospitalization. In community-dwelling elders, lower extremity disability is an important predictor of functional loss. Ambulation assistive devices (canes, walkers), as markers of lower extremity disability, may predict functional decline associated with hospitalization, but this has not been evaluated previously. We sought to determine the association of mobility impairment, as indicated by cane or walker use prehospitalization, with adverse outcomes at hospital discharge and 3 months post discharge. Methods. Subjects were community-dwelling adults (N = 1212) aged 70 and older, hospitalized for acute medical illness. The study was a secondary analysis of the Hospital Outcomes Project for the Elderly, a prospective randomized trial at three university and two private acute-care hospitals, which randomized patients to usual care or an intervention group designed to maintain functional abilities. Results. After controlling for demographic and illness-related characteristics and prehospital function, mobility impairment was significantly associated with functional decline. Use of a walker was associated with 2.8 times increased risk for decline in ADL function by hospital discharge (p = .0002). Three months after discharge, patients who used assistive devices prior to hospitalization were more likely to have declined in both ADLs (p = .02) and IADLs (p = .003). Conclusions. Hospitalized patients with mobility impairment, as indicated by use of a cane or a walker, are at high risk for functional decline. Such patients may benefit from more intensive in-hospital and post-hospital rehabilitative therapy to maintain function .

.ALMOST one third of older adults decline in ability to

~ perform basic activities of daily living (ADLs) related to hospitalization and acute medical illness (1-3). If

function is lost related to acute medical illness, it often is not reversed (3). The ability to predict patients at risk for decline is important for targeting preventive and rehabilitative therapies during and after hospitalization. Risk factors for functional decline with hospitalization include cognitive impairment, pressure sores, low social activity levels, older age, preexisting instrumental ADL impairment, delirium, and specific diseases (3-7). These risk factors indicate that, in general, patients with preexisting impairments are more likely to lose function in the setting of acute illness and hospitalization. In community populations, the presence of lower extremity disability predicts development of new functional dependency (8). Measures of lower extremity disability include objective measures of balance, gait, and chair rise ability (8,9). Use of an ambulation assistive device may also serve as a marker of lower extremity disability. Consistent with this, previous studies have shown that patients who use ambulation aids have poorer performance on objective tests of balance, gait, and mobility (10-12). We hypothesized that use of an ambulation assistive device, as a marker of lower extremity impairment, would predict development of new functional dependency in the setting of acute hospitalization.

We have previously shown that patients who use ambulation aids prior to hospitalization are at increased risk for decline in walking with hospitalization, and for falls after hospital discharge (13,14). However, the association of assistive device use with loss of ADL independence has not been evaluated. The specific objectives of this study were to determine (a) if use of an assistive device prior to hospitalization is associated with increased risk for functional decline during the hospitalization period, independent of decline in walking during the same period; and (b) if degree of risk for decline varies according to type of assistive device (cane vs walker), and (c) if use of an assistive device continues to predict decline in function, compared to prehospital baseline, 3 months after discharge. METHODS

Study populations.-This study is a secondary analysis of the Hospital Outcomes Project for the Elderly (HOPE), a prospective randomized trial at three university and two private acute-care hospitals of distinct but related interventions to prevent functional decline with hospitalization. Subjects eligible to participate in the HOPE study were age 70 or older and hospitalized for a medical illness between 1989 and 1992 at one of five hospitals: Cedars Sinai Medical Center, Los Angeles, CA; University Hospitals of Cleveland, OH; St Mary's Hospital Medical Center, MadiM83

M84

MAHONEYETAL.

son, WI; Stanford University Hospital, Palo Alto, CA; and Yale-New Haven Hospital, New Haven, CT. At each hospital site, patients were randomized to usual care or an intervention designed to maintain functional abilities of hospitalized older persons. The study design, site-specific interventions, and inclusion and exclusion criteria were previously described (3,4,15-.19). Exclusion criteria common to all sites included terminal illness, severe cognitive impairment, inability to give informed consent, admission to the intensive care unit, and residence in a nursing home prior to hospitalization. Informed consent was obtained in accordance with appropriate institutional review boards. A total of 1,486 patients met these eligibility criteria and were enrolled in the study between 1989 and 1992. From this sample, 207 were excluded from analysis for the following reasons: 64 died, and 143 had either absent (87) or incomplete (56) data regarding ADLs at admission or discharge. The remaining 1,279 represented 86% of the eligible study population (90% of survivors). Patients excluded because of absent or incomplete functional information did not differ from the study sample in age, gender, living site, race, marital status, or length of hospital stay. Of the sample of 1,279 subjects, 59 patients reported they used a wheelchair prior to hospitalization and 9 reported use of another mobility aid (non-cane, non-walker). Because these groups were too small to permit meaningful analysis, patients using wheelchairs and other non-cane, non-walker assistive devices were excluded from the analysis. Thus, the population for this study comprised 1,212 subjects who used no assistive device, or a cane or a walker prior to hospitalization.

MeQsurements.-Patients were assessed at three time intervals: within 48 hours of admission, at hospital discharge, and 3 months after discharge. The admission assessment included baseline demographic information (age, gender, race, living arrangement [alone or with others], and marital status), and retrospective self-report of use and type of ambulation assistive device, ability to walk across a small room, and need for help from another person to perform seven IADLs (Lawton scale) and five ADLs (Katz scale) 2 weeks before hospitalization (20,21). Patients were considered independent in walking if they were able to walk across a small room without the assistance of another person. Patients were considered independent in an ADL or IADL if they were able to perform the activity without help from another person. At the Madison site, the Katz scale of discharge ADL function was validated against the Functional Independence Measure (FIM), an objective, performance-based assessment of ADL function (22,23). Overall agreement between the two measures was 78% (17). An abbreviated Mini-Mental State Examination (MMSE), excludirig the 9 items composing the language portion of the test, was also conducted (24). The language items were excluded because of concerns about the ability of acutely ill patients to perform tasks such as writing and copying. Spearman's rank order correlation between the 21-item and standard 30-item MMSE was .90 (p < .001) in the cohort at the Madison site. At discharge, patients were assessed by the same meth-

odology as on admission regarding ability to perform the five ADLs and independently walk across a room. Discharge diagnoses (ICD-9-CM classification), hospital length of stay, and discharge destination (community vs nursing home) were obtained from hospital records. Three months after discharge, patients and/or families were contacted by telephone to determine current ADL, IADL, and walking function, living arrangement, rehospitalization, and mortality.

Data collection.-All data were obtained from participants or proxies by trained interviewers using predetermined protocols and data collection instruments standardized across sites. After informed consent was signed, the participant was identified a priori as the primary source of data, and proxies were used only if the patient was too ill to respond. Participant self-reports accounted for 90% of admission and 85% of discharge information, and 90% of 3-month follow-ups. The same reporter was used for all three phases of data collection in 77% of cases. Analysis.-Type of assistive device used prior to hospitalization was classified as cane or walker. Patients using more than one type-were considered to use the type indicating the greatest degree of impairment, with a walker representing greater impairment than a cane. A summary ADL functional score from 0 to 5 was calculated as the number of ADLs;that a subject could independently perform (out of five basic AD's-bathing, dressing, toileting, transferring and eating). A summary score for IADLs from 0 to 7 was similarly constructed. Logistic regression was performed to identify factors predictive of functional and health utilization outcomes at discharge and at 3 months post discharge. For these analyses, functional outcomes were treated as binary variables, comparing decline in ADL score or decline in IADL score (scored as 1) to no decline in ADL or IADL score (scored as 0). All outcome models tested the independent contributions of demographic characteristics (age, race, gender, living arrangement), prehospitalization ADL and IADL functional status and mobility [use of ambulation assistive device vs no device]), cognitive status (abbreviated MMSE), diagnostic category of principal diagnosis (cancer, circulatory, respiratory, gastrointestinal, and other), and number of comorbid illnesses, controlling for site of hospitalization, treatment group (experimental vs control), and source of data (patient vs other). Loss of walking independence during hospitalization was included as a control variable, to determine if use of an assistive device predicted decline in ADL function independent of change in walking. Loss of walking independence was defined to have occurred if a patient could ambulate across a small room independent of another person prior to hospitalization, but required personal assistance by discharge. To determine the difference in risk associated with cane or walker versus no assistive device, logistic models were repeated using prehospital use of cane and walker as indicator variables, with no assistive device being the reference category. Outcomes of interest at hospital discharge included decline in ADL function compared to baseline two weeks

ASSISTIVE DEVICES AND FUNCTIONAL DECLINE

M85

prior to hospitalization, and discharge to a nursing home. For analysis of decline in function, prehospital ADL function was included in the regression using dummy variables for each level of ADL function prior to hospitalization. Patients with a prehospital ADL level of 0 (n = 29) were excluded because they could not decline. Outcomes of interest at 3 months post discharge included mortality, decline in ADL and IADL function compared to two weeks prior to hospitalization, location of residence (nursing home vs community), and rehospitalization in the interim. For the analysis of decline in ADL function, prehospital ADL function was included as described above, excluding those with a level of 0 (n = 21). Decline in IADL function was analyzed similarly, using indicator variables for prehospital levels of IADL function, and excluding 58 people with a prehospital IADL level of O. Chi-square analysis was performed on the subset of patients who had declined in ADL function by discharge, to compare frequency of subsequent recovery in cane users versus those with no assistive device. Similarly, chi-square testing was performed on the subset of patients who had not declined by hospital discharge, to compare frequency of subsequent ADL decline in cane users versus those with no assistive device. All analyses were performed using SAS statistical software, version 6.09 (Statistical Analysis Systems, Cary, NC).

excluded from analysis of follow-up outcomes. Subjects with complete data at 3 months (n = 1072) did not differ from those with missing data in demographic characteristics, prehospital ADL or IADL function, assistive device use, or cognition. By 3 months post discharge, 17.4% of survivors had declined in ADL function compared to baseline before hospitalization, and 39.3% had declined in IADL function. Seven percent were residing in a nursing home at 3-month follow-up, and 22% had been rehospitalized in the interim. Table 3 shows the association between use of an assistive device prehospitalization and 3-month outcomes. Compared to those who did not use an assistive device, patients who used an ambulation aid were at 1.7 times increased risk for decline in ADL and IADL function, after controlling for decline in walking during the hospitalization period. In examining type of assistive device, both cane and walker were associated with 1.7 times increased risk for ADL decline, although only cane use was statistically significant. The association of cane use with ADL decline was likely related to the fact that cane users were less likely to recover after discharge. Fifty-three percent of cane users

RESULTS

Age (Mean, SD)

79.1 (6.2)

Characteristics of the study sample.-Table 1 shows characteristics of the 1,212 subjects who were admitted to the hospital from the community for acute medical illness and used a cane, walker, or no ambulation aid prior to hospitalization. Subjects were largely independent in ADLs prior to admission, but had substantial impairment in IADLs. One third of subjects used an assistive device prior to hospitalization, with 21% using a cane and 13% a walker. Almost half of patients were admitted for primary respiratory or circulatory conditions; gastrointestinal conditions or cancer accounted for another 20% of admitting diagnoses. Discharge outcomes associated with use of an assistive device.-At hospital discharge, 29.5% of people had declined in one or more of the five ADLs, and 7.8% were discharged to a nursing home. Fifteen percent had become newly dependent on another person to walk. Table 2 shows the risk of adverse outcomes at discharge, in relation to any assistive device use and to type of assistive device, after controlling for decline in walking. Use of a walker prior to hospitalization was associated with 2.8 times increased risk for decline in ADLs (bathing, dressing, toileting, transferring, and eating), independent of prehospital functional status and decline in walking during hospitalization. Use of a cane was not associated with increased risk of ADL decline. Three-month outcomes associated with use of an assistive device.-At 3-month follow-up, 10.3% of subjects had died. Of patients who survived, 6.4% had incomplete information regarding functional status. These patients. were

Table 1. Characteristics of Subjects Prior to Hospitalization (N = 1212)

Female (%)

61.6

Race (%) Caucasian (%) African American (%) Asian (%) Hispanic, American Indian, other (%)

78.8 18.5 2.0 0.2

Living arrangement (%) Alone (%) With spouse (%) With family (%) With unrelated person (%)

37.5 40.8 17.5 4.2

Abbreviated MMSE (range 0-21) (Mean, SD)*

17.1 (3.9)

No. independent ADLs (range 0-5) (Mean, SD)t No. independent IADLs (range 0-7) (Mean, SD):j:

4.5 (1.2) 4.6 (2.2)

Need assistive device before hospitalization Cane (%) Walker(%)

33.9 20.9 13.0

~4

comorbid illnesses (%)

62.3

Discharge diagnostic category: Cancer (%) Gastrointestinal (%) Respiratory (%) Circulatory (%) Other (%)

14.6 19.6 26.2 33.0

6.4

*Abbreviated MMSE = Mini-Mental State Examination, excluding 9 language items. t ADL =5 activities of daily living (bathing, dressing, transferring, toileting, and eating). :j:IADL = 7 instrumental activities of daily living (managing finances, taking medications, telephoning, shopping, using transportation, preparing meals, doing housework).

MAHONEYETAL.

M86

Table 2. Relationship of Prehospital Use of a Cane or a Walker With Functional Decline and Institutionalization at Hospital Discharge

Discharge Outcomes

Adjusted Odds Ratio Associated With Assistive Device*t 95% CI

Adjusted Odds Ratio Associated With Cane*t 95% CI

Adjusted Odds Ratio Associated With Walker*t 95% CI

Decline in ADL function during

hospitalizationjf Discharge to nursing home

1.32 0.83

(0.92-1.92) (0.45-1.52)

0.91 0.62

(0.59-1.4) (0.30-1.28)

2.77 1.25

(1.63-4.72) (0.58-2.69)

*Adjusted for age, race, gender, prehospitalliving arrangement, prehospital ADL and IADL function, cognition, diagnostic category, number of comorbid illnesses, site, treatment group, and source of data. All models adjusted for decline in walking during hospitalization (that is, new dependency on another person to walk across a small room) during index hospitalization. [Odds ratio equals the probability of functional decline if a person uses an assistive device divided by the probability of functional decline if a person does not use an assistive device. tCompared to baseline 2 weeks prior to hospitalization. §Prehospital ADL function included in logistic model with indicator variable for each level of function; patients with prehospital ADL level of 0 were excluded (n = 29).

Table 3. Relationship of Prehospital Use of a Cane or a Walker With Functional Decline and Institutionalization at 3 Months

3-Month Outcomes

Adjusted Odds Ratio Associated With Assistive Device*t 95%CI

Adjusted Odds Ratio Associated With 95%CI Cane*t

Adjusted Odds Ratio Associated With 95%CI Walker*t

Mortality

1.58

(0.95-2.63)

1.54

(0.86-2.75)

1.70

(0.87-3.33)

Among Survivors (n = 1072): Decline in ADL functiontj Decline in IADL functiont** In nursing home Rehospitalized in interim

1.71 1.71 1.19 1.08

(1.10-2.65) (1.19-2.44) (0.59-2.39) (0.73-1.61)

1.72 1.63 0.79 1.20

(1.06-2.81) (1.10-2.42) (0.33-1.87) (0.78-1.85) .

1.72 2.05 1.85 0.91

(0.94-3.14) (1.19-3.52) (0.81-4.21 ) (0.51-1.62)

*Adjusted for age, race, gender, prehospitalliving arrangement, prehospital ADL and IADL function, cognition, diagnostic category, number of comorbid illnesses, site, treatment group, and source of data. All models adjusted for decline in walking (that is, new dependency on another person to walk across a small room) during index hospitalization. tOdds ratio equals the probability of functional decline if a person uses an assistive device divided by the probability of functional decline if a person does not use an assistive device. tCompared to baseline 2 weeks prior to hospitalization. §Prehospital ADL function included in model with indicator variable for each level of function, patients with prehospital ADL level of 0 were excluded from the model (n = 21). **Prehospital IADL function included in model with indicator variable for each level of function, patients with prehospital IADL level of 0 were excluded from the model (n =58).

recovered versus 67% of those who used no ambulation device (p = .04). In addition, compared to those who used no assistive device, cane users were more likely to have new decline in ADL function subsequent to discharge (16% of cane users declined vs 7% of those with no assistive device, p = .001). Prehospital assistive device use also predicted decline in IADL function by 3 months post discharge, with risk being greater for a walker (OR = 2.0, p = .01) than for a cane (OR = 1.6, p = .015). Use of an assistive device prior to hospitalization was not associated with 3-month mortality, nursing home utilization, or rehospitalization. DISCUSSION

In this multicenter study of older adults hospitalized for a medical illness, use of a walker prior to hospitalization was strongly associated with increased risk of functional decline during the hospitalization period. Loss of function was often, but not always, accompanied by loss of walking ability. After controlling for simultaneous loss of walking function, use of a walker continued to be an independent predic-

tor of loss of ADL function. Three months after discharge, use of an assistive device was significantly associated with decline in both ADL and IADL function compared to prehospitalization. Both cane and walker were predictive, with risk in general being higher for a walker than for a cane. Previous studies have shown that older age, impaired cognition, preexisting ADL and IADL impairment, pressure sores, and low social support are risk factors for adverse functional outcomes after hospitalization (3,4,6). The significance of this study is that it demonstrates that specific indicators of mobility impairment are important, additional predictors of functional decline. Our findings suggest that, independent of functional level, preexisting mobility impairment places a patient closer to the threshold of loss of independence in ADLs. The fact that functional decline at discharge was associated with use of a walker but not a cane suggests that the more severe the mobility impairment before hospitalization, the closer the threshold of functional decline. For clinicians, simple questions regarding the use and type of assistive device before admission may permit identification of patients who are at high risk for loss of

ASSISTIVE DEVICES AND FUNCTIONAL DECLINE

ADL independence, and who might benefit from preventive and rehabilitative efforts designed to maintain mobility and ADL function during hospitalization. We have previously shown that use of an assistive device is a significant predictor of loss of walking independence associated with hospitalization for acute medical illness (13). Importantly, in this study, development of new functional dependency related to assistive device use was not just due to loss of walking independence. Use of a walker remained a significant predictor of functional decline even after controlling for loss of walking function. If a patient used a walker prior to hospitalization, he or she remained at risk for loss of independence in other basic ADLs even if walking independence was maintained. Ambulation assistive devices can improve mobility and allow independent performance of mobility-related tasks for people who would otherwise be dependent on others. In one study, half of people using mobility aids reported they could not perform mobility-related activities without the aid (25). In a study by Sonn and Grimby (26), 43% of those who used assistive devices for indoor mobility activities said that the device increased independence, 74% reported that it increased safety, and 79% reported that it was necessary for performance of the activity. In another study, older adult cane users reported that use of a cane improved both functional ability and confidence. (27). Although assistive devices may permit continued independent functioning for patients who would otherwise be dependent on others, at the same time they indicate the presence of substantial impairment in lower extremity function. Use of an assistive device has been associated with decreased balance, strength, mobility, and stair-climbing power, slowed gait speed, and increased risk for falls (10,11,14,28). Fried et al. (12) found that older adults who were independent in ADLs and IADLs, but who modified tasks by use of an assistive device or by other methods, were more impaired on objective tests of physical performance. Use of an assistive device, therefore, would appear to be a marker for a greater degree of physiologic impairment for a given level of ADL or IADL function. As our data suggest, it also serves as a marker for those more likely to become further dependent on others for performance of ADLs. Use of an assistive device before hospitalization also predicted functional decline 3 months later. Patients who used assistive devices were at risk for loss of IADL as well as ADL functions. Reasons for persistent or new decline 3 months after hospitalization are many. Prolonged bedrest, decreased mobility, recurrent illness, and falls may prevent recovery of function after hospitalization. New or recurrent illness may engender new functional dependence over the subsequent 3 months. Further study is needed to determine (a) specific factors that mediate loss of function, or failure to recover function, after hospitalization, and (b) if these factors serve as greater impediments to recovery in patients with preexisting mobility impairment. Use of a cane did not predict decline in ADL function at hospital discharge, relative to no assistive device, but did predict decline 3 months later. Our data suggest two explanations for this. First, use of a cane may be a marker for

M87

decreased ability to recover. Consistent with this, cane users were significantly less likely to recover compared to those who used no assistive device. Second, our data also suggest that use of a cane may be associated with increased risk for new decline in ADL function subsequent to hospitalization among those who had not lost function previously. There are a number of limitations to this analysis. First, determination of functional status was based on subjective report. A previous study (17) has documented that patient self-report may overestimate ADL function compared to objective performance. However, the agreement of our findings at discharge and 3 months later argues for their validity. Second, we could not determine to what extent functional dependence at the time of hospital discharge was related to lack of an assistive device that may have been used at home. However, the continued association of assistive device use with functional decline 3 months later suggests that functional decline at discharge was not due merely to lack of a preexisting assistive device, but to a more severe physiologic decline. In summary, this study suggests that use of an assistive device is an important independent predictor of decline in function around the hospitalization period. Such patients may be easily identified for specific hospital and post-hospital interventions to prevent ADL and IADL dependence. Further study is needed to determine if therapies such as progressive-resistance training, specific balance training, or more intensive use of physical therapy during hospitalization can ameliorate functional decline in this high-risk group (10,29). Further study is also needed to determine if changes in practice style, including more widespread provision of acute care units for the elderly (16), may prevent or reverse functional loss in patients whose preexisting mobility impairment places them at risk. ACKNOWLEDGMENTS

Financial support for this study was provided by Grant 92312-G from the John A. Hartford Foundation, New York, NY. Dr. Mahoney is supported by NIA Academic Award 1 K08 AGOO623-01. This work was also supported in part by the Department of Veterans Affairs. This is Madison VA GRECC publication 97-10. Address correspondence and reprint requests to Dr. Jane Mahoney, GRECC, William S. Middleton Memorial Veterans Hospital, 2500 Overlook Terrace, Madison WI 53705. E-mail: [email protected] REFERENCES

1. Hirsch CH, Sommers L, Olsen A, Mullen L, Winograd CH. The natural history of functional morbidity in hospitalized older patients. JAm Geriatr Soc. 1990;38:1296-1303. 2. McVey LJ, Becker PM, Saltz CC, Feussner JR, Cohen HJ. Effect of a geriatric consultation team on functional status of elderly hospitalized patients. A randomized, controlled clinical study. Ann Intern Med. 1989;110:79-84. 3. Sager MA, Franke T, Inouye SK, et al. Functional outcomes of acute medical illness and hospitalization in older persons. Arch Intern Med. 1996;156:645-652. 4. Sager MA, Rudberg MA, Jalaluddin M, et al. Hospital Admission Risk Profile (HARP): identifying older patients at risk for functional decline following acute medical illness and hospitalization. J Am Geriatr Soc. 1996;44:251-257. 5. Murray AM, Levkoff SE, Wetle TT, et al. Acute delirium and functional decline in the hospitalized elderly patient. J Gerontol Med Sci. 1993;48:MI81-MI86. 6. Inouye SK, Wagner DR, Acampora 0, et al. A predictive index for

M88

7.

8.

9.

10.

11.

12.

13.

14. 15.

16.

17.

MAHONEYETAL.

functional decline in hospitalized elderly medical patients. J Gen Intern Med. 1993;8:645--652. Ferrucci L, Guralnik lM, Pahor M, Corti MC, Havlik RJ. Hospital diagnoses, Medicare charges and nursing home admissions in the year when older persons become severely disabled. JAMA. 1997;277: 728-734. Guralnik lM, Ferrucci L, Simonsick EM, Salive ME, Wallace RB. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332:556-561. Tinetti ME, Inouye SK, Gill TM, Doucette IT. Shared risk factors for falls, incontinence, and functional dependence: unifying the approach to geriatric syndromes. JAMA. 1995;273:1348-1353. Fiatarone MA, O'Neill EF, Ryan ND, et al. Exercise training and nutritional supplementation for physical frailty in very elderly people. N Engl J Med. 1994;330:1769-1775. Berg OK, Maki BE, Williams rr, Holliday PJ, Wood-Dauphinee SL. Clinical and laboratory measures of postural balance in an elderly population. Arch Phys Med Rehabil. 1992;73:1073-1080. Fried LP, Bandeen-Roche K, Williamson JD, et al. Functional decline in older adults: expanding methods of ascertainment. J Gerontol Med Sci. 1996;5IA:M206-M214. Mahoney IE, Sager MA, Jalaluddin M. New walking dependence associated with hospitalization for acute medical illness: incidence and significance. J Gerontol Med Sci. 1998;53A:M307-M312. Mahoney J, Sager M, Dunham NC, Johnson J. Risk of falls after hospital discharge. JAm Geriatr Soc. 1994;42:269-274. Margitic SE, Inouye SK, Thomas JL, Cassel CK, Regenstreif DI, Kowal J. Hospital Outcomes Project for the Elderly (HOPE): Rationale and design for a prospective pooled analysis. J Am Geriatr Soc. 1993;41:258-267. Landefeld CD, Palmer RM, Kresevic DM, Fortinsky RH, Kowal J. A randomized trial of care in a hospital medical unit especially designed to improve the functional outcomes of acutely ill older patients. N Engl J Med. 1995;332:1338-1344. Sager MA, Dunham NC, Schwantes A, Mecum L, Halverson K, Harlowe D. Measurement of activities of daily living in hospitalized elderly: a comparison of self-report and performance-based methods. JAm Geriatr Soc. 1992;40:457-462.

18. Inouye SK, Wagner DR, Acampora D, Horwitz RI, Cooney LM, Tinetti ME. A controlled trial of a nursing-centered intervention in hospitalized elderly medical patients: the Yale Geriatric Care Program. JAm Geriatr Soc. 1993;41:1353-1360. 19. Inouye SK, Acampora D, Miller RL, Fulmer T, Hurst LD, Cooney LM. The Yale Geriatric Care Program: a model of care to prevent functional decline in hospitalized elderly patients. J Am Geriatr Soc. 1993;41: 1345-1352. 20. Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist. 1969;9: 179-186. 21. Katz S, Downs TD, Cash HR, Grotz RC. Progress in development of the index of ADL. Gerontologist. 1970;10:20-30. 22. Hamilton BB, Granger CV, Sherivan FS, et al. A uniform national data system for medical rehabilitation. In: Fuhrer MJ, ed. Rehabilitation Outcomes: Analysis and Management. Baltimore: Paul Brooks Publishing; 1987:137-147. 23. Guide for Use of the Uniform Data Set for Medical Rehabilitation. Buffalo, NY: State University of New York at Buffalo; 1987. 24. Folstein MF, Folstein SE, McHugh PRoMini-Mental State: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189-198. 25. Parker MG, Thorslund M. The use of technical aids among community-based elderly. Am J Occup Ther. 1991;45:712-718. 26. Soon U, Grimby G. Assistive devices in an elderly popultion studied at 70 and 76 years of age. Disabil Rehabil. 1994;16:85-92. 27. Dean E, Ross J. Relationships among cane fitting, function, and falls. Phys Ther. 1993;73:494-504. 28. Wild D, Nayak USL, Isaacs B. Characteristics of old people who fell at home. J Clin Exp Gerontol. 1980;2:271-287. 29. Hu M, Woollacott MH. Multisensory training of standing balance in older adults: I. Postural stability and one-leg stance balance. J Gerontol Med Sci. 1994;49:M52-M61.

Received July 23, 1997 Accepted April 23, 1998

Be a Part of the New VA-VA Central Iowa Health Care System.Knoxville Division The VA Central Iowa Health Care System, a two-division enterprise, is seeking a physician, board certified in Internal Medicine or board certified/board eligible in Geriatric Medicine, for its Knoxville, Iowa, division. This division is a long-term facility accredited by JCAHO that includes an acute psychiatry and CARP-accredited subacute rehabilitation. The incumbent will provide leadership for a 230-bed nursing home, supervision of mid-level practitioners, assist with primary care, and provide consultative services for veterans with medical and/or psychiatric diagnoses. Knoxville, Iowa, is located in a rural setting, 35 miles from Des Moines, and offers all the conveniences of a small town, including a day care center on the medical center grounds, excellent schools, a recreation center, and affordable housing. VA offers a generous benefits package, including 10 paid holidays, federal retirement including a 401k plan, 30 vacation days, and paid sick leave. U.S. citizenship and possession of an active and current license to practice medicine are required. Applicants may forward a current curriculum vitae and letter of intent to VA Central Iowa Health Care System, Knoxville Division, 1515 W. Pleasant, Knoxville, IA 50138. Or contact Usha Jaipaul, MD! Chief, Physical Medicine & Rehabilitation Service at (515) 828-5015. Selectee may be subject to drug testing. VA is an Equal Opportunity Employer.