- Email: [email protected]
Rehabilitation outcomes in cognitively impaired patients admitted to skilled nursing facilities from the community
1602
Rehabilitation Outcomes in Cognitively Impaired Patients Admitted to Skilled Nursing Facilities From the Community Carol Barnes, MS, PT, GCS, Douglas Conner, PhD, Lil Legault, MPT, Nora Reznickova, MD, Cynthia Harrison-Felix, PhD ABSTRACT. Barnes C, Conner D, Legault L, Reznickova N, Harrison-Felix C. Rehabilitation outcomes in cognitively impaired patients admitted to skilled nursing facilities from the community. Arch Phys Med Rehabil 2004;85:1602-7. Objective: To examine the outcomes of patients with varying levels of cognitive impairment who received rehabilitation in skilled nursing facilities (SNFs). Design: A retrospective analysis of the records of people admitted to SNFs for rehabilitation. Setting: Seven SNFs in Colorado. Participants: Community-dwelling persons (N⫽7159), 65 years of age and older, admitted for rehabilitation after a hospitalization or decline in function between May 1998 and May 2002. Interventions: Not applicable. Main Outcome Measures: Cognitive impairment was assessed using a 4-level categorization of the FIM instrument cognitive score at admission. Functional gain was measured using the FIM. Community discharge was measured as the proportion of patients discharged to home, board and care, or assisted living facility. Rehabilitation progress was measured as the number of FIM points gained per day. Results: Significant functional gains were made during rehabilitation in motor and cognitive FIM scores, regardless of cognitive impairment. The most cognitively impaired patients required more rehabilitation intervention, achieved less FIM gain, and were less likely to be discharged to the community. The strongest predictors of FIM gain were the amount of therapy hours and admission cognitive FIM score. The strongest predictors of discharge to the community were the discharge total FIM score and age. The strongest predictors of adequate rehabilitation progress were medical complexity and admission cognitive FIM score. Conclusions: Patients with cognitive impairment were able to recover function with rehabilitation intervention. Patients with a more serious cognitive impairment received more rehabilitation intervention than patients with less impairment. Outcomes were predicted by admission and rehabilitation measures that were qualitatively different from other discharge outcomes. Health care professionals need to consider these
From the Continuing Care Department (Barnes, Reznickova) and Clinical Research Unit (Conner), Kaiser Permanente, Aurora, CO; Department of Physical Therapy, University of Colorado Health Sciences Center, Denver, CO (Barnes); Garden Terrace, Aurora, CO (Legault); and Research Department, Craig Hospital, Englewood, CO (Harrison-Felix). Presented to the American Physical Therapy Association, February 2002, Boston, MA; the American Medical Directors Association, March 2003, Orlando, FL; and the American Geriatric Society, May 2003, Baltimore, MD. No commercial party having a direct interest in the results of the research supporting this article has or will confer a benefit on the author(s) or on any organization with which the author(s) is/are associated. Reprint requests to Carol Barnes, Continuing Care Dept, Kaiser Permanente, 2550 S Parker Rd, Ste 400, Aurora, CO 80014, e-mail: [email protected]. 0003-9993/04/8510-8716$30.00/0 doi:10.1016/j.apmr.2004.02.025
Arch Phys Med Rehabil Vol 85, October 2004
factors as they create a rehabilitation plan of care for patients with cognitive impairment. Key Words: Cognition disorders; Geriatrics; Patient discharge; Rehabilitation © 2004 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation LDERLY PATIENTS WITH DEMENTIA or cognitive E impairment often require time in a skilled nursing facility (SNF) rehabilitation unit to recover functional ability after a hospitalization. Many elderly rehabilitation patients have significant cognitive impairment. As the elderly population continues to grow, the question of the influence of cognitive impairment on rehabilitation outcomes becomes more important. Studies have shown mixed results about the influence of cognition on rehabilitation outcomes. Cognitive impairment can be an important predictor of functional outcome1-16 and discharge destination17-19 in elderly patients. For example, in a study of 204 patients with hip fracture admitted to a hospital geriatric rehabilitation department, impaired cognitive status at admission lowered their rehabilitation outcome.8 As a result of these studies, some health professionals and payer sources have questioned the value of providing rehabilitation services to cognitively impaired patients.12-14,20,21 Other studies22-26 reported comparable functional gains in elderly rehabilitation patients with and without cognitive impairments. For example, Diamond et al24 found, in 52 consecutive admissions to a geriatric rehabilitation unit, that functional improvements in cognitively impaired patients were comparable with cognitively intact patients at the time of discharge. It is not surprising that study outcomes are mixed, given the limited number of diagnoses (often hip fracture or stroke alone), different rehabilitation settings (acute inpatient rehabilitation, geriatric multidisciplinary team rehabilitation approach), small sample sizes, and limited number of facilities. Many studies16,24,25 excluded patients with significant cognitive impairments or other medical conditions. There is little information about the influences on patient outcomes, given that approximately 247,000 patients in SNFs receive rehabilitation services annually.27 Our study examined outcomes in a large population of patients with varying levels of cognitive impairment and different diagnoses in multiple SNFs, using the same outcomes measurement system. Our study was designed to address the following research questions: (1) Is cognitive impairment associated with rehabilitation outcomes? (2) What admission and rehabilitation characteristics predict functional gain and discharge destination, and who will make rehabilitation progress? (3) Can patients with cognitive impairment admitted to SNFs achieve good clinical outcomes with rehabilitation intervention?
1603
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes
METHODS Design Our study was a retrospective analysis of records from patients admitted to SNFs for rehabilitation after a hospitalization or decline in function. All patients were members of a large managed care plan and were accepted for rehabilitation regardless of their cognitive status. The institutional review board of the health plan approved the study. Setting The setting included 7 SNFs in Colorado. SeniorMetrix Inc,28 the database manager for rehabilitation services contracted by the health plan, provided rehabilitation records. Participants Subjects were 7159 community-dwelling individuals, 65 years of age and older, admitted to 1 of 7 different SNFs for rehabilitation between May 1998 and May 2002. They represented a variety of impairment groups, including orthopedic conditions (30%), other disabling conditions (eg, neoplasms, various diseases of the major organ systems, metabolic and immune disorders, injuries, and complications) (24%), pulmonary conditions (10%), stroke (9%), and cardiac conditions (8%). Assessment of Functional Gain Functional ability was measured using the FIM instrument,29,30 which was assessed by treatment providers at the time of SNF admission and at discharge. All providers were trained and credentialed in FIM scoring. The FIM is a valid and reliable measure of functional ability. It consists of a cognitive subscale (5 items), a motor subscale (13 items), and an overall score (all 18 items). Item scores range from 1 (total dependence) to 7 (total independence), resulting in a cognitive FIM score range of 5 to 35 and a motor FIM score range of 13 to 91.31 FIM gain is a commonly used measure of functional improvement that results from rehabilitation intervention. It is calculated as the difference between the admission and discharge FIM and is influenced by the level of function a person has at the time of admission. The higher the FIM score at admission, the less gain can be made before reaching the maximum score. Heruti et al8 accounted for this ceiling effect by using the Montebello Rehabilitation Factor Score. For our study, the raw admission FIM scores were used to preserve the health plan’s measure of adequate progress, which is 1 FIM point per therapy day. The amount of FIM gain possible was limited for only a few patients in our sample; therefore, we believe that the influence of any ceiling effect was minimal.
Measurement of Other Characteristics Age and gender were obtained from rehabilitation records. Other admission characteristics included days postonset and medical complexity. Days postonset is the number of days from the precipitating event to the initiation of rehabilitation treatment. Medical complexity is a clinician-based measure of the effect of a patient’s comorbidities on function developed by SeniorMetrix Inc.28 It is measured on a scale of 0 to 5 (0, no systemic disease other than primary medical diagnosis; 1, inactive and/or irrelevant systemic disease; 2, active, relevant systemic disease not limiting function; 3, active systemic disease limiting function; 4, active systemic disease severely limiting function; 5, moribund or terminal condition). Rehabilitation characteristics included the number of therapy hours (physical, occupational, and speech therapies combined) and cycle length of stay (LOS; number of therapy days). Data Analysis Statistical analyses were performed using SAS statistical software.32,a The 3 outcomes considered were the FIM gain, the proportion of patients discharged to the community, and the proportion of patients making adequate progress. Significant mean differences among the 4 cognitive impairment groups for continuous variables were tested using analysis of variance techniques. Post hoc mean differences were tested using the Scheffé multiple comparisons procedure. Adjustments were made for nonnormally distributed data. Significant differences among the 4 impairment groups for categoric variables were analyzed using chi-square tests. Multivariate linear regression techniques were used to identify predictors of FIM gain. All admission and rehabilitation characteristics were entered into each model as independent variables using a forward stepwise selection technique. Multivariate logistic regression was used to identify predictors of discharge disposition and for achieving the health plan’s measure of adequate rehabilitation progress. A forward stepwise selection technique was used. Total FIM score and FIM gain were entered into the initial logistic regression model only for the discharge to the community outcome because total FIM at discharge and FIM gain are closely associated with adequate progress (the numerator for adequate progress is FIM gain). Odds ratios (ORs) and 95% confidence intervals (CIs) were determined from each of the final models. ORs are also presented for a 5-unit increase for each continuous variable (when possible) by raising the OR to the fifth power. The number of therapy hours and cycle LOS were entered separately into all regression models because there was a strong association between these 2 variables (r⫽.77). For each outcome, the variable contributing to the stronger model is reported based on the Akaike information criterion coefficient. RESULTS
Discharge Destination The number of patients discharged to the community was measured as those discharged to home, board and care, or assisted living facility. Assessment of Cognitive Impairment Cognitive impairment was assessed using a 4-level categorization of the FIM cognitive score at admission: group 1, needing more than 50% assistance (score of ⬍3 on any cognitive FIM item); group 2, needing less than 50% assistance (score of 3 or 4 on any cognitive FIM item); group 3, requiring supervision only (score of 5 on any cognitive FIM item); and group 4, independent (score of ⬎5 on any cognitive FIM item).
Patient Characteristics The average age was 80.4⫾7.7 years (range, 65–107y). Women comprised 66.9% of the sample. Table 1 includes patient and rehabilitation admission characteristics by severity of cognitive impairment. Patients with more severe cognitive impairment had lower admission FIM and motor FIM scores, greater medical complexity, and more days postonset and were more likely to be older and men. Cognitive Impairment and Rehabilitation Outcomes Rehabilitation and outcome characteristics for the cognitive impairment categories are shown in table 2. Patients with the Arch Phys Med Rehabil Vol 85, October 2004
1604
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes Table 1: Patient and Rehabilitation Admission Characteristics Among Cognitive Impairment Categories Cognitive Impairment Category Characteristic
Independent (n⫽1676)
Supervision (n⫽3871)
⬍50% Assistance (n⫽1383)
⬎50% Assistance (n⫽229)
P Value*
Age (y) Days postonset Admission FIM score Admission motor FIM score Admission cognitive FIM score Medical complexity储 Gender (% women)
78.5⫾7.4 7.2⫾12.9 85.9⫾10.8 52.1⫾10.6 33.8⫾1.5 2.6⫾1.0 70.6
80.8⫾7.8 8.0⫾14.1 69.1⫾14.1 43.8⫾11.8 25.3⫾4.7 3.0⫾.80 65.9
81.7⫾7.6 8.7⫾16.5 58.7⫾17.1 38.7⫾12.9 19.9⫾6.6 3.1⫾.83 66.6
80.8⫾7.9 12.6⫾24.9 29.9⫾14.6 22.1⫾11.4 7.8⫾4.5 3.3⫾.85 59.3
⬍.001† ⬍.001‡ ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001
NOTE. Values are mean ⫾ standard deviation (SD) unless otherwise noted. *Analysis of variance (ANOVA) for continuous measures with Scheffé multiple comparison tests and chi-square test for proportions. All contrasts among levels of cognitive impairment differed significantly unless otherwise noted. † Significant contrasts: independent vs ⬍50% and ⬎50% assistance. ‡ Significant contrasts: independent vs each other category. 储 Larger value equals greater medical complexity.
greatest cognitive impairments (requiring assistance) needed significantly more therapy hours than did independent patients or those patients requiring only supervision. Patients needing more than 50% assistance had a significantly longer cycle LOS and were discharged with significantly less FIM gain. Fewer of these patients made a minimum of 1 FIM point per therapy day (adequate progress), even though those patients achieved an average of 1.2 FIM points per day. They were also significantly less likely to be discharged to the community. Predictors of Rehabilitation Outcomes Table 3 shows the results of the multiple regression analysis identifying predictors of FIM gain and admission and rehabilitation characteristics. Increasing therapy hours were strongly associated with increasing FIM gain, followed by higher admission cognitive FIM score, lower admission motor FIM score, and decreasing medical complexity. Younger age and fewer days postonset were less strongly associated with FIM gain. Gender was not associated with FIM gain. These significant characteristics accounted for about 13% of the variance in FIM gain (overall r2⫽13.4), which suggests that there were other important predictors of FIM gain that were not accounted for in this analysis. Results of the logistic regression analysis of the likelihood of a patient being discharged to the community are shown in table 4. Total discharge FIM score was the strongest predictor of
discharge to the community. A 1-point increase in the discharge FIM score increased the likelihood of community discharge by 4.8% (OR⫽1.048; 95% CI, 1.042–1.055). A 5-point increase in discharge FIM score increased the probability of community discharge nearly 25%. A 1-year increase in age reduced the likelihood of community discharge by 3.5% (OR⫽.965; 95% CI, .957–.973), whereas a 5-year increase in age decreased the likelihood of community discharge by 16%. A 1-point total FIM gain increased the probability of community discharge by 2.4% (OR⫽.976; 95% CI, .967–.984), whereas a 5-point FIM gain improved the probability of community discharge to 12.6%. A 1-day increase in the number of days of therapy reduced the likelihood of community discharge by 2.4%, whereas a 5-day increase in cycle LOS decreased the probability of community discharge to 11.4%. Fewer days postonset also contributed to the likelihood of community discharge but to a lesser degree (OR⫽.995; 95% CI, .991– .998). Results of the logistic regression analysis of the likelihood of achieving the health plan’s criterion for adequate progress of 1 FIM point per therapy day are shown in table 5. A 1-unit increase in medical complexity decreased the likelihood of achieving adequate progress by 22.1% (OR⫽.779; 95% CI, .719 –.845). An increase of 1 point in the admission cognitive FIM score increased the likelihood of achieving adequate progress by 5.2%, whereas an increase of 5 points increased the
Table 2: Rehabilitation and Outcome Characteristics Among Cognitive Impairment Categories Cognitive Impairment Category Characteristic
Independent (n⫽1676)
Supervision (n⫽3871)
⬍50% Assistance (n⫽1383)
⬎50% Assistance (n⫽229)
P Value*
Cycle LOS (d) Total therapy hours Total FIM gain Cognitive FIM gain Motor FIM gain Total FIM points/days of therapy Discharged to the community (%) Adequate progress (%)
8.6⫾5.7 13.1⫾8.8 20.6⫾11.7 ⫺.23⫾2.3 20.8⫾10.7 3.3⫾3.1 89.1 89.1
11.1⫾7.6 16.8⫾12.2 20.6⫾13.8 1.3⫾3.0 19.3⫾12 2.4⫾2.6 75.1 79.4
12.0⫾8.6 18.3⫾14.5 20.0⫾15.8 2.0⫾4.0 18.0⫾13.2 2.0⫾2.7 64.4 72.7
14.3⫾12.6 20.3⫾19.7 14.9⫾17.5 3.2⫾5.0 11.7⫾13.7 1.2⫾1.7 44.1 45.0
⬍.001 ⬍.001† ⬍.001‡ ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001
NOTE. Values are mean ⫾ SD unless otherwise noted. *ANOVA for continuous measures with Scheffé multiple comparison tests and chi-square test for proportions. All contrasts among levels of cognitive impairment differed significantly unless otherwise noted. † Significant contrasts: all comparisons except between ⬍50% and ⬎50% assistance. ‡ Significant contrasts: ⬎50% assistance vs ⬍50% assistance, supervision, and independent.
Arch Phys Med Rehabil Vol 85, October 2004
1605
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes Table 3: Multiple Linear Regression Analysis of Significant Admission and Rehabilitation Predictors of FIM Gain*
Predictor
Standardized Parameter Estimate
Standard Error
P Value
Therapy hours Cognitive FIM score at admission Motor FIM score at admission Complexity Age Days postonset
.285 .177 ⫺.124 ⫺.018 ⫺.083 ⫺.066
.013 .026 .015 .187 .020 .010
⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001
Table 5: Logistic Regression: Adjusted OR and 95% CI for Admission and Treatment Predictors of Adequate Rehabilitation Progress* Predictor
*r 2⫽.134.
Change
Medical complexity Admission cognitive FIM score Cycle LOS Admission motor FIM score Age Days postonset
OR
95% CI
1-unit increase
0.779
0.719–0.845
1-point increase 1-d increase
1.052 0.951
1.041–1.062 0.944–0.959
1-point increase 1-y increase 1-d increase
1.020 0.986 0.989
1.014–1.026 0.978–0.994 0.985–0.992
*Adequate rehabilitation progress is defined as 1 FIM point per day of therapy.
likelihood to 28.8%. A 1-day increase in cycle LOS decreased the probability of making adequate progress by almost 5% (OR⫽.951; 95% CI, .944 –.959), whereas a 5-day increase in their stay reduced the likelihood of making adequate progress by 22.2%. A single-point increase in admission motor FIM score increased the probability of adequate progress by 2% (OR⫽1.020; 95% CI, 1.014 –1.026), whereas a 5-point increase raised this to 10.4%. A 1-year increase in age decreased a patient’s chance of achieving adequate progress by 1.4% (OR⫽.986; 95% CI, .978 –.994), whereas a 5-year increase in age decreased the chance of making adequate progress to 6.8%. Finally, a 1-day increase in days postonset resulted in a 1.1% decrease in the likelihood of making adequate progress (OR⫽.989; 95% CI, .985–.992), whereas an increase of 5 days reduced a patient’s chances of making adequate progress to 5.4%. DISCUSSION Postacute geriatric rehabilitation in SNFs will grow in importance as the population ages and more people require this service. These patients have a variety of diagnoses and variable cognitive abilities. Rehabilitation professionals are responsible for developing a plan of care for these patients and need a greater understanding of how cognitive impairment affects rehabilitation outcomes. Regardless of the level of cognitive impairment, most patients achieved good clinical outcomes for the 3 outcomes that were measured: FIM gain, adequate rehabilitation progress, and discharge to the community. However, the importance of cognitive impairment, as well as other admission and rehabilitation measures, depends on the outcome being measured. Admission and Rehabilitation Measures Affecting Different Outcomes FIM gain and what is defined by the health plan as adequate rehabilitation progress are closely related because adequate progress is simply the FIM gain divided by the number of days of therapy (cycle LOS). Admission cognitive and motor FIM
Table 4: Logistic Regression: Adjusted OR and 95% CI for Predictors of Discharge to the Community Predictor
Change
OR
95% CI
Discharge total FIM score Age FIM gain at discharge Cycle LOS Days postonset
1-point increase 1-y increase 1-point increase 1-d increase 1-d increase
1.048 0.965 1.024 0.976 0.995
1.042–1.055 0.957–0.973 1.016–1.032 0.967–0.984 0.991–0.998
scores were both strong predictors of FIM gain and adequate progress. Admission motor FIM score was inversely associated with FIM gain but directly associated with adequate progress. Discharge to the community represents an outcome different than either FIM gain or adequate rehabilitation progress. For patients to be discharged to the community, they must have a certain level of function (represented by a minimum total FIM score at discharge) that is sufficient to allow them to function in a community setting; they may also need to satisfy other requirements that are independent of rehabilitation, such as having a place to live or a caregiver in the home.33 The amount of functional gain that a patient achieves is less important than reaching a certain threshold level of function. The strongest predictor of community discharge was a patient’s discharge FIM score. There may be some point at which higher discharge FIM scores significantly increase the likelihood of community discharge. Studies have found that FIM scores above a certain level increase the likelihood of discharge to home or to a community setting.34,35 Younger age was the second strongest predictor of community discharge but to a lesser degree. Increasing FIM gain increased the probability of community discharge about the same as cycle LOS (table 4). Severe Cognitive Impairment Does Not Preclude Rehabilitation Our analyses suggest that cognitive impairment is an important influence on FIM gain and adequate rehabilitation progress, but it is not as important in determining community discharge. Poor cognitive status does not, however, preclude good clinical outcomes. Even though the most severe cognitively impaired group was significantly older and more medically complex, almost half returned to the community, averaged a 14.9 FIM point gain (vs 20.6 in the intact group), and averaged a 1.2 FIM point gain per therapy day during their rehabilitation (0.2 points per day greater than the health plan’s benchmark of adequate progress). Diamond et al24 found that greater cognitive deficits were associated with greater nursing home placement, but they also found that 38% of subjects with severe cognitive impairment and most patients with mild to moderate impairment returned home after discharge. Implications for Rehabilitation Efforts Patients in the most severe cognitively impaired group required a longer LOS and more therapy hours to obtain a good clinical outcome. Our results and those of others22-26 support the inclusion of cognitively impaired older adults in rehabilitation programs. Rehabilitation providers should realize that persons with poor cognitive status will often require longer and/or more specialized36 rehabilitation efforts to achieve maxArch Phys Med Rehabil Vol 85, October 2004
1606
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes
imal rehabilitation gains. Cognitive impairment does not, as a rule, prevent good outcomes. Decisions about who will benefit from rehabilitation services need to be made considering all aspects of a patient’s clinical picture. Clinicians need to address cognitive impairments on a case-by-case basis in the plan of care. The goal of rehabilitation also needs to be considered. Our study looked at predictive admission and rehabilitation characteristics for 3 rehabilitation outcomes and found that different characteristics were important for different outcomes. Predicting community discharge is important to plan for the assistance a person may need at home or what alternative care setting would be appropriate. FIM gain is important in assessing rehabilitation potential and response to therapy intervention. Adequate progress helps a clinician determine whether the patient meets the Medicare qualifications for an SNF level of care. Limitations This was a retrospective study of a large number of SNF patients 65 years of age and older. By the nature of the large sample size, statistical significance may be found without corresponding clinical significance. However, it was thought that the findings from our study were relevant and clinically significant. All patients were members of a group model managed care plan that uses a geriatric rehabilitation model with their contracted facilities, providing on-site physician services and care coordination. Because of this, most patients older than 65 years were admitted to an SNF for rehabilitation rather than to an acute rehabilitation hospital facility. The results of our study may not be applicable to all SNF rehabilitation programs. Community discharge was defined as discharge to home, board and care, or assisted living. Results of this outcome are limited by how it is defined. Our study did not look at other variables that may influence a patient’s ability to return to the community, such as the amount of family support.33 Our study did not look at the role of cognitive impairment for different admission diagnoses, which one would expect to have a strong influence on the results presented here. Finally, our analysis did not look at individual scale items making up cognitive and motor FIM scores and their relative importance in predicting rehabilitation outcomes.37,38 All these areas are important to consider in future research. CONCLUSIONS The objective of our study was to examine the discharge outcomes of patients with different levels of cognitive impairment who received rehabilitation in SNFs. The results suggest that, regardless of the severity of cognitive impairment, patients were able to recover significant function with rehabilitation intervention. Patients with more serious cognitive impairments required more rehabilitation intervention than patients with less impairment. Each rehabilitation outcome was predicted by admission and rehabilitation measures that were qualitatively different from other discharge outcomes. Health care professionals need to consider these factors as they create a rehabilitation plan of care for patients with cognitive impairment. Acknowledgments: Chris Wirtalla, SeniorMetrix Inc, Nashville, TN, provided the rehabilitation database and assisted the investigators with the measures collected. The following facilities collected the data used in this study: Boulder Manor, Boulder, CO; Mariner Health of Arch Phys Med Rehabil Vol 85, October 2004
Denver, Denver, CO; Cherry Creek Nursing Center, Aurora, CO; Western Hills Health Care Center, Lakewood, CO; Garden TerraceColorado, Aurora, CO; Life Care Center of Westminster, Westminster, CO; and Life Care Center of Longmont, Longmont, CO. References 1. Kong KH, Chua KS, Tow AP. Clinical characteristics and functional outcome of stroke patients 75 years old and older. Arch Phys Med Rehabil 1998;79:1535-9. 2. Paolucci S, Antonucci G, Gialloreti LE, et al. Predicting stroke inpatient rehabilitation outcome: the prominent role of neuropsychological disorders. Eur Neurol 1996;36:385-90. 3. Ozdemir F, Birtane M, Tabatabaei R, Ekuklu G, Kokino S. Cognitive evaluation and functional outcome after stroke. Am J Phys Med Rehabil 2001;80:410-5. 4. Galski T, Bruno RL, Zorowitz R, Walker J. Predicting length of stay, functional outcome and aftercare in the rehabilitation of stroke patients. Stroke 1993;24:1794-800. Comment in: Stroke 1994;25:1295. 5. Lichtenberg PA, MacNeill SE, Mast BT. Environmental press and adaptation to disability in hospitalized live-alone older adults. Gerontologist 2000;40:549-56. 6. Patel MD, Coshall C, Rudd AG, Wolfe CD. Cognitive impairment after stroke: clinical determinants and its associations with longterm stroke outcomes. J Am Geriatr Soc 2002;50:700-6. 7. Moroney JT, Desmond DW. Factors predictive of stroke outcome in a rehabilitative setting. Neurology 1997;48:1475-7. 8. Heruti RJ, Lusky A, Barell V, Ohry A, Adunsky A. Cognitive status at admission: does it affect the rehabilitation outcome of elderly patients with hip fracture? Arch Phys Med Rehabil 1999; 80:432-6. 9. Cardoc-Davies TH. Medical profiles of patients admitted to a geriatric assessment and rehabilitation unit. N Z Med J 1987;100: 557-9. 10. Lehmann JF, de Lateur BJ, Fowler Jr RS, et al. Stroke rehabilitation: outcome and prediction. Arch Phys Med Rehabil 1975;56: 383-9. 11. Rubenstein LZ, Wieland D, English P, Josephson K, Sayre JA, Abrass IB. The Sepulveda VA Geriatric Evaluation Unit: data on four-year outcomes and predictors of improved patient outcomes. J Am Geriatr Soc 1984;32:503-12. 12. Billig N, Ahmed SW, Kenmore PI. Hip fracture, depression, and cognitive impairment: a follow-up study. Orthop Rev 1988;17: 315-20. 13. Magaziner J, Simonsick EM, Kashner TM, Hebel JR, Kenzora JE. Predictors of functional recovery one year following hospital discharge for hip fracture: a prospective study. J Gerontol 1990; 45:M101-7. 14. Landi F, Bernabei R, Russo A, et al. Predictors of rehabilitation outcomes in frail patients treated in a geriatric hospital. J Am Geriatr Soc 2002;50:679-84. 15. Marcantonio ER, Flacker JM, Michaels M, Resnick NM. Delirium is independently associated with poor functional recovery after hip fracture. J Am Geriatr Soc 2000;48:618-24. 16. Ruchinskas RA, Singer HK, Repetz NK. Clock drawing, clock copying, and physical abilities in geriatric rehabilitation. Arch Phys Med Rehabil 2001;82:920-4. 17. Friedman PJ, Baskett JJ, Richmond DE. Cognitive impairment and its relationship to gait rehabilitation in the elderly. N Z Med J 1989;102:603-6. 18. Walsh PF. Long-term community outcome following medical rehabilitation in older adults [abstract]. Dissertation Abstracts Int 2000;61(1-B):551. 19. Mokler PJ, Sandstrom R, Griffin M, Farris L, Jones C. Predicting discharge destination for patients with severe motor stroke: im-
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes
20.
21. 22. 23.
24.
25.
26.
27. 28. 29.
portant functional tasks. Neurorehabil Neural Repair 2000;14:181-5. Cummings SR, Phillips SL, Wheat ME, et al. Recovery of function after hip fracture. The role of social supports. J Am Geriatr Soc 1988;36:801-6. Resnick B, Daly MP. The effect of cognitive status on outcomes following rehabilitation. Fam Med 1997;29:400-5. Barnes B. Ambulation outcomes after hip fracture. Phys Ther 1984;64:317-20. Davies AD, Gargaro P, Dodd A, Smith C. Mental status and occupational therapy outcome in teaching dressing skills to the disabled elderly. Clin Rehabil 1989;3:293-8. Diamond PT, Felsenthal G, Macciocchi SN, Butler DH, LallyCassady D. Effect of cognitive impairment on rehabilitation outcome. Am J Phys Med Rehabil 1996;75:40-3. Goldstein FC, Strasser DC, Woodard JL, Roberts VJ. Functional outcomes of cognitively impaired hip fracture patients on a geriatric rehabilitation unit. J Am Geriatr Soc 1997;45:35-42. Comment in: J Am Geriatr Soc 1997;45:1406-7. Cartwright DL, Madill HM, Dennis S. Cognitive impairment and functional performance of patients admitted to a geriatric assessment and rehabilitation centre. Phys Occup Ther Geriatr 1996; 14(3):1-21. Facts and trends 2001: nursing facility sourcebook. Washington (DC): American Health Care Association; 2001. Postacute rehabilitation [database]. Nashville: SeniorMetrix Inc; 1998. Granger CV, Hamilton BB, Keith RA, Zielezny M, Sherwin FS. Advances in functional assessment for medical rehabilitation. Top Geriatr Rehabil 1986;1(3):59-74.
1607
30. Uniform Data System for Medical Rehabilitation. The FIM system. Available at: http://www.udsmr.org. Accessed March 2, 2004. 31. Guide for the Uniform Data Set for Medical Rehabilitation (adult FIM). Version 4.0. Buffalo (NY): Uniform Data System for Medical Rehabilitation, UB Foundation Activities Inc; 1993. 32. SAS Institute Inc. SAS/STAT user’s guide, version 6. 4th ed. Vol 2. Cary: SAS Institute Inc; 1989. 33. Ween JE, Alexander MP, D’Esposito M, Roberts M. Factors predictive of stroke outcome in a rehabilitation setting. Neurology 1996;47:388-92. 34. Kosasih JB, Borca HH, Wenninger WJ, Duthie E. Nursing home rehabilitation after acute rehabilitation: predictors of outcomes. Arch Phys Med Rehabil 1998;79:670-3. 35. Black TM, Soltis T, Bartlett C. Using the Functional Independence Measure instrument to predict stroke rehabilitation outcomes. Rehabil Nurs 1999;24(3):109-14. 36. Huusko TM, Karppi P, Avikainen V, Kautiainen H, Sulkava R. Randomized, clinically controlled trial of intensive geriatric rehabilitation in patients with hip fracture: subgroup analysis of patients with dementia. BMJ 2000;321:1107-11. 37. Mauthe RW, Haaf DC, Hayn P, Krall JM. Predicting discharge destination of stroke patients using a mathematical model based on six items from the Functional Independence Measure. Arch Phys Med Rehabil 1996;77:10-3. 38. Stineman MG, Maislin G, Fiedler RC, Granger CV. A prediction model for functional recovery in stroke. Stroke 1997;28:550-6. Supplier a. SAS Institute Inc, 100 SAS Campus Dr, Cary, NC 27513.
Arch Phys Med Rehabil Vol 85, October 2004
Rehabilitation Outcomes in Cognitively Impaired Patients Admitted to Skilled Nursing Facilities From the Community Carol Barnes, MS, PT, GCS, Douglas Conner, PhD, Lil Legault, MPT, Nora Reznickova, MD, Cynthia Harrison-Felix, PhD ABSTRACT. Barnes C, Conner D, Legault L, Reznickova N, Harrison-Felix C. Rehabilitation outcomes in cognitively impaired patients admitted to skilled nursing facilities from the community. Arch Phys Med Rehabil 2004;85:1602-7. Objective: To examine the outcomes of patients with varying levels of cognitive impairment who received rehabilitation in skilled nursing facilities (SNFs). Design: A retrospective analysis of the records of people admitted to SNFs for rehabilitation. Setting: Seven SNFs in Colorado. Participants: Community-dwelling persons (N⫽7159), 65 years of age and older, admitted for rehabilitation after a hospitalization or decline in function between May 1998 and May 2002. Interventions: Not applicable. Main Outcome Measures: Cognitive impairment was assessed using a 4-level categorization of the FIM instrument cognitive score at admission. Functional gain was measured using the FIM. Community discharge was measured as the proportion of patients discharged to home, board and care, or assisted living facility. Rehabilitation progress was measured as the number of FIM points gained per day. Results: Significant functional gains were made during rehabilitation in motor and cognitive FIM scores, regardless of cognitive impairment. The most cognitively impaired patients required more rehabilitation intervention, achieved less FIM gain, and were less likely to be discharged to the community. The strongest predictors of FIM gain were the amount of therapy hours and admission cognitive FIM score. The strongest predictors of discharge to the community were the discharge total FIM score and age. The strongest predictors of adequate rehabilitation progress were medical complexity and admission cognitive FIM score. Conclusions: Patients with cognitive impairment were able to recover function with rehabilitation intervention. Patients with a more serious cognitive impairment received more rehabilitation intervention than patients with less impairment. Outcomes were predicted by admission and rehabilitation measures that were qualitatively different from other discharge outcomes. Health care professionals need to consider these
From the Continuing Care Department (Barnes, Reznickova) and Clinical Research Unit (Conner), Kaiser Permanente, Aurora, CO; Department of Physical Therapy, University of Colorado Health Sciences Center, Denver, CO (Barnes); Garden Terrace, Aurora, CO (Legault); and Research Department, Craig Hospital, Englewood, CO (Harrison-Felix). Presented to the American Physical Therapy Association, February 2002, Boston, MA; the American Medical Directors Association, March 2003, Orlando, FL; and the American Geriatric Society, May 2003, Baltimore, MD. No commercial party having a direct interest in the results of the research supporting this article has or will confer a benefit on the author(s) or on any organization with which the author(s) is/are associated. Reprint requests to Carol Barnes, Continuing Care Dept, Kaiser Permanente, 2550 S Parker Rd, Ste 400, Aurora, CO 80014, e-mail: [email protected]. 0003-9993/04/8510-8716$30.00/0 doi:10.1016/j.apmr.2004.02.025
Arch Phys Med Rehabil Vol 85, October 2004
factors as they create a rehabilitation plan of care for patients with cognitive impairment. Key Words: Cognition disorders; Geriatrics; Patient discharge; Rehabilitation © 2004 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation LDERLY PATIENTS WITH DEMENTIA or cognitive E impairment often require time in a skilled nursing facility (SNF) rehabilitation unit to recover functional ability after a hospitalization. Many elderly rehabilitation patients have significant cognitive impairment. As the elderly population continues to grow, the question of the influence of cognitive impairment on rehabilitation outcomes becomes more important. Studies have shown mixed results about the influence of cognition on rehabilitation outcomes. Cognitive impairment can be an important predictor of functional outcome1-16 and discharge destination17-19 in elderly patients. For example, in a study of 204 patients with hip fracture admitted to a hospital geriatric rehabilitation department, impaired cognitive status at admission lowered their rehabilitation outcome.8 As a result of these studies, some health professionals and payer sources have questioned the value of providing rehabilitation services to cognitively impaired patients.12-14,20,21 Other studies22-26 reported comparable functional gains in elderly rehabilitation patients with and without cognitive impairments. For example, Diamond et al24 found, in 52 consecutive admissions to a geriatric rehabilitation unit, that functional improvements in cognitively impaired patients were comparable with cognitively intact patients at the time of discharge. It is not surprising that study outcomes are mixed, given the limited number of diagnoses (often hip fracture or stroke alone), different rehabilitation settings (acute inpatient rehabilitation, geriatric multidisciplinary team rehabilitation approach), small sample sizes, and limited number of facilities. Many studies16,24,25 excluded patients with significant cognitive impairments or other medical conditions. There is little information about the influences on patient outcomes, given that approximately 247,000 patients in SNFs receive rehabilitation services annually.27 Our study examined outcomes in a large population of patients with varying levels of cognitive impairment and different diagnoses in multiple SNFs, using the same outcomes measurement system. Our study was designed to address the following research questions: (1) Is cognitive impairment associated with rehabilitation outcomes? (2) What admission and rehabilitation characteristics predict functional gain and discharge destination, and who will make rehabilitation progress? (3) Can patients with cognitive impairment admitted to SNFs achieve good clinical outcomes with rehabilitation intervention?
1603
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes
METHODS Design Our study was a retrospective analysis of records from patients admitted to SNFs for rehabilitation after a hospitalization or decline in function. All patients were members of a large managed care plan and were accepted for rehabilitation regardless of their cognitive status. The institutional review board of the health plan approved the study. Setting The setting included 7 SNFs in Colorado. SeniorMetrix Inc,28 the database manager for rehabilitation services contracted by the health plan, provided rehabilitation records. Participants Subjects were 7159 community-dwelling individuals, 65 years of age and older, admitted to 1 of 7 different SNFs for rehabilitation between May 1998 and May 2002. They represented a variety of impairment groups, including orthopedic conditions (30%), other disabling conditions (eg, neoplasms, various diseases of the major organ systems, metabolic and immune disorders, injuries, and complications) (24%), pulmonary conditions (10%), stroke (9%), and cardiac conditions (8%). Assessment of Functional Gain Functional ability was measured using the FIM instrument,29,30 which was assessed by treatment providers at the time of SNF admission and at discharge. All providers were trained and credentialed in FIM scoring. The FIM is a valid and reliable measure of functional ability. It consists of a cognitive subscale (5 items), a motor subscale (13 items), and an overall score (all 18 items). Item scores range from 1 (total dependence) to 7 (total independence), resulting in a cognitive FIM score range of 5 to 35 and a motor FIM score range of 13 to 91.31 FIM gain is a commonly used measure of functional improvement that results from rehabilitation intervention. It is calculated as the difference between the admission and discharge FIM and is influenced by the level of function a person has at the time of admission. The higher the FIM score at admission, the less gain can be made before reaching the maximum score. Heruti et al8 accounted for this ceiling effect by using the Montebello Rehabilitation Factor Score. For our study, the raw admission FIM scores were used to preserve the health plan’s measure of adequate progress, which is 1 FIM point per therapy day. The amount of FIM gain possible was limited for only a few patients in our sample; therefore, we believe that the influence of any ceiling effect was minimal.
Measurement of Other Characteristics Age and gender were obtained from rehabilitation records. Other admission characteristics included days postonset and medical complexity. Days postonset is the number of days from the precipitating event to the initiation of rehabilitation treatment. Medical complexity is a clinician-based measure of the effect of a patient’s comorbidities on function developed by SeniorMetrix Inc.28 It is measured on a scale of 0 to 5 (0, no systemic disease other than primary medical diagnosis; 1, inactive and/or irrelevant systemic disease; 2, active, relevant systemic disease not limiting function; 3, active systemic disease limiting function; 4, active systemic disease severely limiting function; 5, moribund or terminal condition). Rehabilitation characteristics included the number of therapy hours (physical, occupational, and speech therapies combined) and cycle length of stay (LOS; number of therapy days). Data Analysis Statistical analyses were performed using SAS statistical software.32,a The 3 outcomes considered were the FIM gain, the proportion of patients discharged to the community, and the proportion of patients making adequate progress. Significant mean differences among the 4 cognitive impairment groups for continuous variables were tested using analysis of variance techniques. Post hoc mean differences were tested using the Scheffé multiple comparisons procedure. Adjustments were made for nonnormally distributed data. Significant differences among the 4 impairment groups for categoric variables were analyzed using chi-square tests. Multivariate linear regression techniques were used to identify predictors of FIM gain. All admission and rehabilitation characteristics were entered into each model as independent variables using a forward stepwise selection technique. Multivariate logistic regression was used to identify predictors of discharge disposition and for achieving the health plan’s measure of adequate rehabilitation progress. A forward stepwise selection technique was used. Total FIM score and FIM gain were entered into the initial logistic regression model only for the discharge to the community outcome because total FIM at discharge and FIM gain are closely associated with adequate progress (the numerator for adequate progress is FIM gain). Odds ratios (ORs) and 95% confidence intervals (CIs) were determined from each of the final models. ORs are also presented for a 5-unit increase for each continuous variable (when possible) by raising the OR to the fifth power. The number of therapy hours and cycle LOS were entered separately into all regression models because there was a strong association between these 2 variables (r⫽.77). For each outcome, the variable contributing to the stronger model is reported based on the Akaike information criterion coefficient. RESULTS
Discharge Destination The number of patients discharged to the community was measured as those discharged to home, board and care, or assisted living facility. Assessment of Cognitive Impairment Cognitive impairment was assessed using a 4-level categorization of the FIM cognitive score at admission: group 1, needing more than 50% assistance (score of ⬍3 on any cognitive FIM item); group 2, needing less than 50% assistance (score of 3 or 4 on any cognitive FIM item); group 3, requiring supervision only (score of 5 on any cognitive FIM item); and group 4, independent (score of ⬎5 on any cognitive FIM item).
Patient Characteristics The average age was 80.4⫾7.7 years (range, 65–107y). Women comprised 66.9% of the sample. Table 1 includes patient and rehabilitation admission characteristics by severity of cognitive impairment. Patients with more severe cognitive impairment had lower admission FIM and motor FIM scores, greater medical complexity, and more days postonset and were more likely to be older and men. Cognitive Impairment and Rehabilitation Outcomes Rehabilitation and outcome characteristics for the cognitive impairment categories are shown in table 2. Patients with the Arch Phys Med Rehabil Vol 85, October 2004
1604
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes Table 1: Patient and Rehabilitation Admission Characteristics Among Cognitive Impairment Categories Cognitive Impairment Category Characteristic
Independent (n⫽1676)
Supervision (n⫽3871)
⬍50% Assistance (n⫽1383)
⬎50% Assistance (n⫽229)
P Value*
Age (y) Days postonset Admission FIM score Admission motor FIM score Admission cognitive FIM score Medical complexity储 Gender (% women)
78.5⫾7.4 7.2⫾12.9 85.9⫾10.8 52.1⫾10.6 33.8⫾1.5 2.6⫾1.0 70.6
80.8⫾7.8 8.0⫾14.1 69.1⫾14.1 43.8⫾11.8 25.3⫾4.7 3.0⫾.80 65.9
81.7⫾7.6 8.7⫾16.5 58.7⫾17.1 38.7⫾12.9 19.9⫾6.6 3.1⫾.83 66.6
80.8⫾7.9 12.6⫾24.9 29.9⫾14.6 22.1⫾11.4 7.8⫾4.5 3.3⫾.85 59.3
⬍.001† ⬍.001‡ ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001
NOTE. Values are mean ⫾ standard deviation (SD) unless otherwise noted. *Analysis of variance (ANOVA) for continuous measures with Scheffé multiple comparison tests and chi-square test for proportions. All contrasts among levels of cognitive impairment differed significantly unless otherwise noted. † Significant contrasts: independent vs ⬍50% and ⬎50% assistance. ‡ Significant contrasts: independent vs each other category. 储 Larger value equals greater medical complexity.
greatest cognitive impairments (requiring assistance) needed significantly more therapy hours than did independent patients or those patients requiring only supervision. Patients needing more than 50% assistance had a significantly longer cycle LOS and were discharged with significantly less FIM gain. Fewer of these patients made a minimum of 1 FIM point per therapy day (adequate progress), even though those patients achieved an average of 1.2 FIM points per day. They were also significantly less likely to be discharged to the community. Predictors of Rehabilitation Outcomes Table 3 shows the results of the multiple regression analysis identifying predictors of FIM gain and admission and rehabilitation characteristics. Increasing therapy hours were strongly associated with increasing FIM gain, followed by higher admission cognitive FIM score, lower admission motor FIM score, and decreasing medical complexity. Younger age and fewer days postonset were less strongly associated with FIM gain. Gender was not associated with FIM gain. These significant characteristics accounted for about 13% of the variance in FIM gain (overall r2⫽13.4), which suggests that there were other important predictors of FIM gain that were not accounted for in this analysis. Results of the logistic regression analysis of the likelihood of a patient being discharged to the community are shown in table 4. Total discharge FIM score was the strongest predictor of
discharge to the community. A 1-point increase in the discharge FIM score increased the likelihood of community discharge by 4.8% (OR⫽1.048; 95% CI, 1.042–1.055). A 5-point increase in discharge FIM score increased the probability of community discharge nearly 25%. A 1-year increase in age reduced the likelihood of community discharge by 3.5% (OR⫽.965; 95% CI, .957–.973), whereas a 5-year increase in age decreased the likelihood of community discharge by 16%. A 1-point total FIM gain increased the probability of community discharge by 2.4% (OR⫽.976; 95% CI, .967–.984), whereas a 5-point FIM gain improved the probability of community discharge to 12.6%. A 1-day increase in the number of days of therapy reduced the likelihood of community discharge by 2.4%, whereas a 5-day increase in cycle LOS decreased the probability of community discharge to 11.4%. Fewer days postonset also contributed to the likelihood of community discharge but to a lesser degree (OR⫽.995; 95% CI, .991– .998). Results of the logistic regression analysis of the likelihood of achieving the health plan’s criterion for adequate progress of 1 FIM point per therapy day are shown in table 5. A 1-unit increase in medical complexity decreased the likelihood of achieving adequate progress by 22.1% (OR⫽.779; 95% CI, .719 –.845). An increase of 1 point in the admission cognitive FIM score increased the likelihood of achieving adequate progress by 5.2%, whereas an increase of 5 points increased the
Table 2: Rehabilitation and Outcome Characteristics Among Cognitive Impairment Categories Cognitive Impairment Category Characteristic
Independent (n⫽1676)
Supervision (n⫽3871)
⬍50% Assistance (n⫽1383)
⬎50% Assistance (n⫽229)
P Value*
Cycle LOS (d) Total therapy hours Total FIM gain Cognitive FIM gain Motor FIM gain Total FIM points/days of therapy Discharged to the community (%) Adequate progress (%)
8.6⫾5.7 13.1⫾8.8 20.6⫾11.7 ⫺.23⫾2.3 20.8⫾10.7 3.3⫾3.1 89.1 89.1
11.1⫾7.6 16.8⫾12.2 20.6⫾13.8 1.3⫾3.0 19.3⫾12 2.4⫾2.6 75.1 79.4
12.0⫾8.6 18.3⫾14.5 20.0⫾15.8 2.0⫾4.0 18.0⫾13.2 2.0⫾2.7 64.4 72.7
14.3⫾12.6 20.3⫾19.7 14.9⫾17.5 3.2⫾5.0 11.7⫾13.7 1.2⫾1.7 44.1 45.0
⬍.001 ⬍.001† ⬍.001‡ ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001
NOTE. Values are mean ⫾ SD unless otherwise noted. *ANOVA for continuous measures with Scheffé multiple comparison tests and chi-square test for proportions. All contrasts among levels of cognitive impairment differed significantly unless otherwise noted. † Significant contrasts: all comparisons except between ⬍50% and ⬎50% assistance. ‡ Significant contrasts: ⬎50% assistance vs ⬍50% assistance, supervision, and independent.
Arch Phys Med Rehabil Vol 85, October 2004
1605
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes Table 3: Multiple Linear Regression Analysis of Significant Admission and Rehabilitation Predictors of FIM Gain*
Predictor
Standardized Parameter Estimate
Standard Error
P Value
Therapy hours Cognitive FIM score at admission Motor FIM score at admission Complexity Age Days postonset
.285 .177 ⫺.124 ⫺.018 ⫺.083 ⫺.066
.013 .026 .015 .187 .020 .010
⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001
Table 5: Logistic Regression: Adjusted OR and 95% CI for Admission and Treatment Predictors of Adequate Rehabilitation Progress* Predictor
*r 2⫽.134.
Change
Medical complexity Admission cognitive FIM score Cycle LOS Admission motor FIM score Age Days postonset
OR
95% CI
1-unit increase
0.779
0.719–0.845
1-point increase 1-d increase
1.052 0.951
1.041–1.062 0.944–0.959
1-point increase 1-y increase 1-d increase
1.020 0.986 0.989
1.014–1.026 0.978–0.994 0.985–0.992
*Adequate rehabilitation progress is defined as 1 FIM point per day of therapy.
likelihood to 28.8%. A 1-day increase in cycle LOS decreased the probability of making adequate progress by almost 5% (OR⫽.951; 95% CI, .944 –.959), whereas a 5-day increase in their stay reduced the likelihood of making adequate progress by 22.2%. A single-point increase in admission motor FIM score increased the probability of adequate progress by 2% (OR⫽1.020; 95% CI, 1.014 –1.026), whereas a 5-point increase raised this to 10.4%. A 1-year increase in age decreased a patient’s chance of achieving adequate progress by 1.4% (OR⫽.986; 95% CI, .978 –.994), whereas a 5-year increase in age decreased the chance of making adequate progress to 6.8%. Finally, a 1-day increase in days postonset resulted in a 1.1% decrease in the likelihood of making adequate progress (OR⫽.989; 95% CI, .985–.992), whereas an increase of 5 days reduced a patient’s chances of making adequate progress to 5.4%. DISCUSSION Postacute geriatric rehabilitation in SNFs will grow in importance as the population ages and more people require this service. These patients have a variety of diagnoses and variable cognitive abilities. Rehabilitation professionals are responsible for developing a plan of care for these patients and need a greater understanding of how cognitive impairment affects rehabilitation outcomes. Regardless of the level of cognitive impairment, most patients achieved good clinical outcomes for the 3 outcomes that were measured: FIM gain, adequate rehabilitation progress, and discharge to the community. However, the importance of cognitive impairment, as well as other admission and rehabilitation measures, depends on the outcome being measured. Admission and Rehabilitation Measures Affecting Different Outcomes FIM gain and what is defined by the health plan as adequate rehabilitation progress are closely related because adequate progress is simply the FIM gain divided by the number of days of therapy (cycle LOS). Admission cognitive and motor FIM
Table 4: Logistic Regression: Adjusted OR and 95% CI for Predictors of Discharge to the Community Predictor
Change
OR
95% CI
Discharge total FIM score Age FIM gain at discharge Cycle LOS Days postonset
1-point increase 1-y increase 1-point increase 1-d increase 1-d increase
1.048 0.965 1.024 0.976 0.995
1.042–1.055 0.957–0.973 1.016–1.032 0.967–0.984 0.991–0.998
scores were both strong predictors of FIM gain and adequate progress. Admission motor FIM score was inversely associated with FIM gain but directly associated with adequate progress. Discharge to the community represents an outcome different than either FIM gain or adequate rehabilitation progress. For patients to be discharged to the community, they must have a certain level of function (represented by a minimum total FIM score at discharge) that is sufficient to allow them to function in a community setting; they may also need to satisfy other requirements that are independent of rehabilitation, such as having a place to live or a caregiver in the home.33 The amount of functional gain that a patient achieves is less important than reaching a certain threshold level of function. The strongest predictor of community discharge was a patient’s discharge FIM score. There may be some point at which higher discharge FIM scores significantly increase the likelihood of community discharge. Studies have found that FIM scores above a certain level increase the likelihood of discharge to home or to a community setting.34,35 Younger age was the second strongest predictor of community discharge but to a lesser degree. Increasing FIM gain increased the probability of community discharge about the same as cycle LOS (table 4). Severe Cognitive Impairment Does Not Preclude Rehabilitation Our analyses suggest that cognitive impairment is an important influence on FIM gain and adequate rehabilitation progress, but it is not as important in determining community discharge. Poor cognitive status does not, however, preclude good clinical outcomes. Even though the most severe cognitively impaired group was significantly older and more medically complex, almost half returned to the community, averaged a 14.9 FIM point gain (vs 20.6 in the intact group), and averaged a 1.2 FIM point gain per therapy day during their rehabilitation (0.2 points per day greater than the health plan’s benchmark of adequate progress). Diamond et al24 found that greater cognitive deficits were associated with greater nursing home placement, but they also found that 38% of subjects with severe cognitive impairment and most patients with mild to moderate impairment returned home after discharge. Implications for Rehabilitation Efforts Patients in the most severe cognitively impaired group required a longer LOS and more therapy hours to obtain a good clinical outcome. Our results and those of others22-26 support the inclusion of cognitively impaired older adults in rehabilitation programs. Rehabilitation providers should realize that persons with poor cognitive status will often require longer and/or more specialized36 rehabilitation efforts to achieve maxArch Phys Med Rehabil Vol 85, October 2004
1606
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes
imal rehabilitation gains. Cognitive impairment does not, as a rule, prevent good outcomes. Decisions about who will benefit from rehabilitation services need to be made considering all aspects of a patient’s clinical picture. Clinicians need to address cognitive impairments on a case-by-case basis in the plan of care. The goal of rehabilitation also needs to be considered. Our study looked at predictive admission and rehabilitation characteristics for 3 rehabilitation outcomes and found that different characteristics were important for different outcomes. Predicting community discharge is important to plan for the assistance a person may need at home or what alternative care setting would be appropriate. FIM gain is important in assessing rehabilitation potential and response to therapy intervention. Adequate progress helps a clinician determine whether the patient meets the Medicare qualifications for an SNF level of care. Limitations This was a retrospective study of a large number of SNF patients 65 years of age and older. By the nature of the large sample size, statistical significance may be found without corresponding clinical significance. However, it was thought that the findings from our study were relevant and clinically significant. All patients were members of a group model managed care plan that uses a geriatric rehabilitation model with their contracted facilities, providing on-site physician services and care coordination. Because of this, most patients older than 65 years were admitted to an SNF for rehabilitation rather than to an acute rehabilitation hospital facility. The results of our study may not be applicable to all SNF rehabilitation programs. Community discharge was defined as discharge to home, board and care, or assisted living. Results of this outcome are limited by how it is defined. Our study did not look at other variables that may influence a patient’s ability to return to the community, such as the amount of family support.33 Our study did not look at the role of cognitive impairment for different admission diagnoses, which one would expect to have a strong influence on the results presented here. Finally, our analysis did not look at individual scale items making up cognitive and motor FIM scores and their relative importance in predicting rehabilitation outcomes.37,38 All these areas are important to consider in future research. CONCLUSIONS The objective of our study was to examine the discharge outcomes of patients with different levels of cognitive impairment who received rehabilitation in SNFs. The results suggest that, regardless of the severity of cognitive impairment, patients were able to recover significant function with rehabilitation intervention. Patients with more serious cognitive impairments required more rehabilitation intervention than patients with less impairment. Each rehabilitation outcome was predicted by admission and rehabilitation measures that were qualitatively different from other discharge outcomes. Health care professionals need to consider these factors as they create a rehabilitation plan of care for patients with cognitive impairment. Acknowledgments: Chris Wirtalla, SeniorMetrix Inc, Nashville, TN, provided the rehabilitation database and assisted the investigators with the measures collected. The following facilities collected the data used in this study: Boulder Manor, Boulder, CO; Mariner Health of Arch Phys Med Rehabil Vol 85, October 2004
Denver, Denver, CO; Cherry Creek Nursing Center, Aurora, CO; Western Hills Health Care Center, Lakewood, CO; Garden TerraceColorado, Aurora, CO; Life Care Center of Westminster, Westminster, CO; and Life Care Center of Longmont, Longmont, CO. References 1. Kong KH, Chua KS, Tow AP. Clinical characteristics and functional outcome of stroke patients 75 years old and older. Arch Phys Med Rehabil 1998;79:1535-9. 2. Paolucci S, Antonucci G, Gialloreti LE, et al. Predicting stroke inpatient rehabilitation outcome: the prominent role of neuropsychological disorders. Eur Neurol 1996;36:385-90. 3. Ozdemir F, Birtane M, Tabatabaei R, Ekuklu G, Kokino S. Cognitive evaluation and functional outcome after stroke. Am J Phys Med Rehabil 2001;80:410-5. 4. Galski T, Bruno RL, Zorowitz R, Walker J. Predicting length of stay, functional outcome and aftercare in the rehabilitation of stroke patients. Stroke 1993;24:1794-800. Comment in: Stroke 1994;25:1295. 5. Lichtenberg PA, MacNeill SE, Mast BT. Environmental press and adaptation to disability in hospitalized live-alone older adults. Gerontologist 2000;40:549-56. 6. Patel MD, Coshall C, Rudd AG, Wolfe CD. Cognitive impairment after stroke: clinical determinants and its associations with longterm stroke outcomes. J Am Geriatr Soc 2002;50:700-6. 7. Moroney JT, Desmond DW. Factors predictive of stroke outcome in a rehabilitative setting. Neurology 1997;48:1475-7. 8. Heruti RJ, Lusky A, Barell V, Ohry A, Adunsky A. Cognitive status at admission: does it affect the rehabilitation outcome of elderly patients with hip fracture? Arch Phys Med Rehabil 1999; 80:432-6. 9. Cardoc-Davies TH. Medical profiles of patients admitted to a geriatric assessment and rehabilitation unit. N Z Med J 1987;100: 557-9. 10. Lehmann JF, de Lateur BJ, Fowler Jr RS, et al. Stroke rehabilitation: outcome and prediction. Arch Phys Med Rehabil 1975;56: 383-9. 11. Rubenstein LZ, Wieland D, English P, Josephson K, Sayre JA, Abrass IB. The Sepulveda VA Geriatric Evaluation Unit: data on four-year outcomes and predictors of improved patient outcomes. J Am Geriatr Soc 1984;32:503-12. 12. Billig N, Ahmed SW, Kenmore PI. Hip fracture, depression, and cognitive impairment: a follow-up study. Orthop Rev 1988;17: 315-20. 13. Magaziner J, Simonsick EM, Kashner TM, Hebel JR, Kenzora JE. Predictors of functional recovery one year following hospital discharge for hip fracture: a prospective study. J Gerontol 1990; 45:M101-7. 14. Landi F, Bernabei R, Russo A, et al. Predictors of rehabilitation outcomes in frail patients treated in a geriatric hospital. J Am Geriatr Soc 2002;50:679-84. 15. Marcantonio ER, Flacker JM, Michaels M, Resnick NM. Delirium is independently associated with poor functional recovery after hip fracture. J Am Geriatr Soc 2000;48:618-24. 16. Ruchinskas RA, Singer HK, Repetz NK. Clock drawing, clock copying, and physical abilities in geriatric rehabilitation. Arch Phys Med Rehabil 2001;82:920-4. 17. Friedman PJ, Baskett JJ, Richmond DE. Cognitive impairment and its relationship to gait rehabilitation in the elderly. N Z Med J 1989;102:603-6. 18. Walsh PF. Long-term community outcome following medical rehabilitation in older adults [abstract]. Dissertation Abstracts Int 2000;61(1-B):551. 19. Mokler PJ, Sandstrom R, Griffin M, Farris L, Jones C. Predicting discharge destination for patients with severe motor stroke: im-
OUTCOMES IN THE COGNITIVELY IMPAIRED, Barnes
20.
21. 22. 23.
24.
25.
26.
27. 28. 29.
portant functional tasks. Neurorehabil Neural Repair 2000;14:181-5. Cummings SR, Phillips SL, Wheat ME, et al. Recovery of function after hip fracture. The role of social supports. J Am Geriatr Soc 1988;36:801-6. Resnick B, Daly MP. The effect of cognitive status on outcomes following rehabilitation. Fam Med 1997;29:400-5. Barnes B. Ambulation outcomes after hip fracture. Phys Ther 1984;64:317-20. Davies AD, Gargaro P, Dodd A, Smith C. Mental status and occupational therapy outcome in teaching dressing skills to the disabled elderly. Clin Rehabil 1989;3:293-8. Diamond PT, Felsenthal G, Macciocchi SN, Butler DH, LallyCassady D. Effect of cognitive impairment on rehabilitation outcome. Am J Phys Med Rehabil 1996;75:40-3. Goldstein FC, Strasser DC, Woodard JL, Roberts VJ. Functional outcomes of cognitively impaired hip fracture patients on a geriatric rehabilitation unit. J Am Geriatr Soc 1997;45:35-42. Comment in: J Am Geriatr Soc 1997;45:1406-7. Cartwright DL, Madill HM, Dennis S. Cognitive impairment and functional performance of patients admitted to a geriatric assessment and rehabilitation centre. Phys Occup Ther Geriatr 1996; 14(3):1-21. Facts and trends 2001: nursing facility sourcebook. Washington (DC): American Health Care Association; 2001. Postacute rehabilitation [database]. Nashville: SeniorMetrix Inc; 1998. Granger CV, Hamilton BB, Keith RA, Zielezny M, Sherwin FS. Advances in functional assessment for medical rehabilitation. Top Geriatr Rehabil 1986;1(3):59-74.
1607
30. Uniform Data System for Medical Rehabilitation. The FIM system. Available at: http://www.udsmr.org. Accessed March 2, 2004. 31. Guide for the Uniform Data Set for Medical Rehabilitation (adult FIM). Version 4.0. Buffalo (NY): Uniform Data System for Medical Rehabilitation, UB Foundation Activities Inc; 1993. 32. SAS Institute Inc. SAS/STAT user’s guide, version 6. 4th ed. Vol 2. Cary: SAS Institute Inc; 1989. 33. Ween JE, Alexander MP, D’Esposito M, Roberts M. Factors predictive of stroke outcome in a rehabilitation setting. Neurology 1996;47:388-92. 34. Kosasih JB, Borca HH, Wenninger WJ, Duthie E. Nursing home rehabilitation after acute rehabilitation: predictors of outcomes. Arch Phys Med Rehabil 1998;79:670-3. 35. Black TM, Soltis T, Bartlett C. Using the Functional Independence Measure instrument to predict stroke rehabilitation outcomes. Rehabil Nurs 1999;24(3):109-14. 36. Huusko TM, Karppi P, Avikainen V, Kautiainen H, Sulkava R. Randomized, clinically controlled trial of intensive geriatric rehabilitation in patients with hip fracture: subgroup analysis of patients with dementia. BMJ 2000;321:1107-11. 37. Mauthe RW, Haaf DC, Hayn P, Krall JM. Predicting discharge destination of stroke patients using a mathematical model based on six items from the Functional Independence Measure. Arch Phys Med Rehabil 1996;77:10-3. 38. Stineman MG, Maislin G, Fiedler RC, Granger CV. A prediction model for functional recovery in stroke. Stroke 1997;28:550-6. Supplier a. SAS Institute Inc, 100 SAS Campus Dr, Cary, NC 27513.
Arch Phys Med Rehabil Vol 85, October 2004