Clinical question: does the literature indicate that patients with a stroke have better outcomes after receiving rehabilitation from an acute rehabilitation facility than from a skilled nursing facility?

The purpose of "Evidence in Practice" is to illustrate the literature search process to obtain evidence that can guide clinical decision making. This article is not a case report. The examination, evaluation, and intervention sections are purposely abbreviated.

A 76-year-old, right-hand dominant man awoke with slurred speech and left-sided weakness and was admitted to the acute hospital facility where I (DKM) was completing a clinical education requirement for my doctor of physical therapy degree. The patient was a retired guitar player who lived at home with his wife. Before his hospitalization, the patient was independent in activities of daily living (ADLs) and community ambulation without an assistive device, and he enjoyed an active lifestyle, including taking daily walks with his wife, gardening, and playing his guitar. Evidence of a recent lacunar infarction in the right internal capsule was found on a computed tomography (CT) scan, and he was diagnosed with an embolic stroke. After 2 days of medical care (including aspirin and heparin), his condition had stabilized, and he was referred to the physical therapy department by the attending neurologist for examination, evaluation, and discharge recommendations.

This patient's medical history included transient ischemic attacks, hypertension, hypercholesterolemia, and a previous left cerebrovascular accident (CVA)--a thalamic infarct--that was noted in the most recent radiology report. He did not experience any residual effects from that CVA and was discharged home from acute care. No rehabilitation services were recommended, and he was independent with all mobility and ADLs following the CVA. The patient had a history of glaucoma, cataracts, and benign prostatic hypertrophy. He stated that he quit smoking 5 years ago (50 packs/year) and that he did not drink large amounts of alcohol.

The patient was alert and oriented to person and time and was able to follow 1- and 2-step commands. He had mild dysarthria, no evidence of expressive or receptive aphasia, and no evidence of left-side neglect. At rest, his vital signs were within normal limits with the exception of blood pressure, which was 130/80 on hypertensive medication. His hemodynamic responses to activity were reassessed after examining bed mobility and were found to be appropriate.

The patient was nonambulatory. He required maximum assistance from 1 person to complete the task of rolling to the right. He could initiate rolling to the left without assistance, but required moderate assistance from 1 person to complete the roll. To rise from a supine position to a sitting position on the left side of his bed, he required maximum assistance from 1 person. With right upper-extremity support, the patient was able to sit at the edge of the bed independently, but trunk weakness kept him from maintaining his balance in midline for more than 30 seconds without upper-extremity support. Sit-to-stand transfers required moderate assistance because he was unable to support his weight through his left lower extremity, and he required moderate assistance from 1 person to maintain standing.

Muscle force in the right upper and lower extremities was within functional limits based on Kendall et al. (1) The patient had no active movement in his left hand or wrist, and only a trace of movement at his elbow. When he was asked to raise his left upper extremity, scapular elevation occurred immediately. He could isolate movement at the shoulder through partial range of motion when gravity was minimized. He was able to initiate left hip flexion against gravity and extend the left knee from 90 degrees to 70 degrees of flexion. He had no active movement in the left ankle. No impairments were noted in passive range of motion or in sensory integrity.

The CT scan suggested that the size of the infarct was relatively small; however, because of the concentration of descending motor fibers in the internal capsule, the patient had moderate to severe strength deficits. People who have had a CVA with purely motor system deficits often reach higher functional levels compared with those with motor and sensory deficits or with motor, sensory, and visual deficits.) This patient had only motor system involvement, whereas his cognition, sensation, perception, and vision were preserved. I determined that he had a good prognosis for improvement in functional status and that he was a good candidate for rehabilitation.

Upon completion of my initial examination, I discussed discharge possibilities with my clinical instructor and then with the patient's physician and case manager. The patient wanted to participate in physical therapy and rehabilitation with the hope of returning home with his wile. They lived in a single-story home, with one step to enter, and they had 2 adult sons living in the area. His wife was in good health and fully independent and thus was capable of providing care at home for him if needed. She recognized the severity of his current limitations, however, and agreed with the rehabilitation team that he was unsafe to return directly home at the time. We discussed discharge possibilities, including acute rehabilitation and skilled nursing facilities (SNFs), and the case manager presented both options for the patient The family inquired about transferring the patient to a SNF close to their home; however, I was inclined to recommend that this patient be discharged to an acute rehabilitation facility where he would receive more intensive therapy compared with that provided by a SNF.

I had no prior experience in the acute care setting in making a discharge recommendation for a patient after a stroke. In order to make the most appropriate discharge recommendation for this patient, I wanted to know whether the type of rehabilitation setting influenced outcomes in patients who have had a stroke. Because of the current emphasis on cost containment at my facility, discharge to the least expensive facility that provided effective recovery for the patient was encouraged. Although I wanted to ensure that I was being fiscally responsible to all parties involved in the rehabilitation of this patient, I also wanted to provide a recommendation based on what I thought was the best health care advice available. I was not familiar with the literature comparing the effectiveness of rehabilitation in an acute rehabilitation setting to rehabilitation at a SNF, so I decided to search the literature.

* Database used for search: CINAHL

The Cumulative Index to Nursing and Allied Health (CINAHL) (www.cinahl.com) * is a database of more than 1,600 journals related to nursing and allied health that is updated monthly. I chose this database as an alternative to the popular MEDLINE database because CINAHL indexes more journals related to allied health professions than MEDLINE. (3) As I later discovered during my search, the indexing terms in CINAHL (analogous to the Medical Subject Heading [MESH] terms in MEDLINE) also provide greater specificity related to physical therapy. For example, the common terms for "acute rehabilitation" in MEDLINE and CINAHL are: "brain injury," "spinal cord injury," "middle aged," "aged," "rehabilitation centers," and "rehabilitation." CINAHL supplements those terms with: "clinical assessment tools," "cerebrovascular accident," and "functional status." Because functional outcomes for patients with CVAs who have received rehabilitation were my primary interest, I thought this database would provide a more efficient search of the literature for tbis case. I accessed CINAHL through my university's subscription to OVID (www.ovid. corn) ([dagger]) on August 29, 2004.

* Keywords for initial search: stroke, acute rehabilitation, and skilled nursing facility

Above the keyword box on the main search page is a check box labeled Map Terms to Subject Headings. I was unsure of its function, so I unchecked the box before beginning my search. I then typed stroke in the keyword box, intending to combine this search and a second search using acute rehabilitation as the keywords (Table: search lines 1 and 2). Next, I typed 1 and 2 in the keyword box to combine the search lines from my first 2 searches, which returned 51 references. I scanned the titles for those that seemed to address rehabilitation settings related to outcomes for patients after stroke and found 4 that appeared to be relevant to my question (Fig. 1). I clicked the box next to each citation to save them (Table: search line 4). Most of the titles of the other articles indicated that they were about medical therapies, particular interventions, or other topics unrelated to nay search. In a fourth search, I used skilled nursing facility as my keyword. As I had done previously with stroke and acute rehabilitation, I combined the search of skilled nursing facility (the fifth search line) with my initial search on stroke (the first search line) by typing 1 and 5 into the query box (Table: search line 6). This resulted in 5 references.

I scanned the titles, again focusing on rehabilitation settings related to poststroke outcomes; however, this revealed only one relevant article, which was the same article by Keith et al (Fig. 1: citation 2) found by my search on stroke and acute rehabilitation. I read the abstracts for all 4 articles by clicking on the Abstract link below each citation. After reading the abstracts, I found that only 2 articles (Fig. 1: citations 1 and 2) were appropriate for nay question. I retrieved the full text of these articles at my university's library. I was dissatisfied, however, with the small number of articles identified with this first search attempt; consequently, I decided to change nay search strategy.

* Keywords for revised search: acute rehabilitation and skilled nursing facility

My main goal was to find literature comparing (1) outcomes for patients with stroke who had received rehabilitation in an acute rehabilitation facility with (2) outcomes for patients with stroke who had received rehabilitation in a SNEI, therefore, combined the terms acute rehabilitation and skilled nursing facility instead of combining each term individually with stroke. I also explored the subject headings (roughly equivalent to MEDLINE's MeSH terms), choosing to leave the Map Terms to Subject Heading box checked, rather than searching for my terms only as keywords. The difference between using subject headings and searching for terms as keywords is that the keyword option limits the search by the specific keyword or phrase in the title, abstract, or text of the article, whereas the subject heading allowed me to find all articles related to the subject heading as well as those that included similar terms (eg, stroke and cerebral vascular accident versus stroke only). I considered switching to the MEDLINE database as well; however, as mentioned above, I found that the subject headings in CINAHL were more specific to the information I was attempting to retrieve. It is possible to combine and simultaneously search multiple databases when using OVID; however, this eliminates the possibility, of using the Map Terms to Subject Headings checkbox. Because I had decided to try my search with subject headings, I chose to continue with a single database.

I retyped acute rehabilitation in the keyword box and clicked Perform Search, which brought me to a Mapping Display page with subject headings (Fig. 2). One of the subject headings under acute rehabilitation was "cerebral vascular accident." Without using "stroke" as a keyword, I was still able to incorporate the topic into my search using the subject heading. To the far fight of each subject heading is a Scope button that connects to a page outlining the scope of the subject heading. I used this tool to confirm that the subject headings rehabilitation reenters, cerebral vascular accident, rehabilitation, and functional status were relevant to nay search. On the Mapping Display page, I checked the box next to each of those subject headings to include them in my search (Fig. 2).

[FIGURE 2 OMITTED]

I also had the option to "explode" or "focus" my search from the Mapping Display page (Fig. 2). Exploding the search would allow me to expand my search to include all of the related terms for the respective subject beading (which can be found by clicking on the subject heading). Focusing the search narrows the search to articles that have the specific subject heading as the focus. I chose not to use the Focus/ Explode feature to (1) avoid reflexing articles outside the scope of my search and (2) avoid limiting my search as I had done previously by using only keywords.

Remaining on the Mapping Display page, I chose to combine my selections by selecting the Boolean operator "OR" from the dropdown list at the top right of the page, so that my search would result in articles addressing any or all of the subject headings. (Choosing "AND" would only return articles that included all of my subject headings.) I then clicked on (continue to resume my search, which resulted in 16,513 references (Table: search line 7). I completed the same procedure for skilled nursing facility. For this phrase, skilled nursing facilities was the only subject heading offered. I read the scope (Fig. 3), and again I decided against either exploding or focusing the search, thus I clicked on Previous Page to go back. After clicking Continue from the Mapping Display page, I was taken to a Subheading Display page (Fig. 4), which displayed subheadings such as "economics," "evaluation," and "utilization." Because I did not want to restrict the focus of my search, I clicked the box Include All Subheadings, and kept the Boolean operator as "OR." I clicked on Continue, and the search produced 907 references (Table: search line 8). At this point, I was ready to combine these 2 searches to locate literature combining acute rehabilitation and skilled nursing facilities. I typed 7 and 8 in the keyword box, which returned 93 citations (Table: search line 9). I scanned the titles of these citations for articles that seemed appropriate for my clinical question, and I found 9 additional articles for review, one of which I had already found in the first search (Fig. 5: citation 1). Again, I was looking for titles that suggested the article made some form of comparison between different rehabilitation settings and effect on outcome.

[FIGURES 3-4 OMITTED]

* Selection of articles for review: I read the abstracts, when available, for each of these articles. When an abstract link was not available, clicking on Bibliographic Links took me to the PubMed Web site (www.ncbi.nlm.nih.gov/PubMed) and the abstract. An abstract was unavailable for the Frohlich and Fogehnan article (Fig. 5: citation 9). Unfortunately, my university's library did not subscribe to this journal, so I excluded this article from my search. I could have retrieved the article through interlibrary loan; however, this was not an option because it typically takes 2 or 3 weeks to receive an article and I had only 1 or 2 days before the patient would be transferred. After reading the abstracts from citations 2, 4, and 5 (Fig. 5), 1 determined that they were not appropriate for nay clinical question (eg, the studies did not compare the 2 settings, did not discuss outcomes related to rehabilitation setting). The first citation listed in Figure 5 was the duplicate from my first search, for which I already had the full-text article. Based on the abstracts, the remaining 4 articles (Fig. 5: citations 3, 6-8) were relevant, and the full text of these articles was available to me, either through Ovid or through the library.

I obtained the full text of the articles from both searches and read the articles to see if they addressed my topic specifically enough. I then used the CAT (critically appraised topic) format as described by Fetters et al (4) to examine those articles that were specific to my question. According to Fetters et al, (4) a CAT is a standardized one-page summary of a research article "organized around a clinical question." A CAT provides a critique of the evidence (based on the methods used and on statistical criteria such as internal, external, and statistical validly') and a statement about the clinical relevance of the results. The citations for the 5 articles are listed in Figure 6. I evaluated the threats to, and the strengths associated with, internal, external, and statistical validity in each article. This process enabled me to formulate a clinical bottom line--that is, the clinical actions I would take--based on my clinical question and the applicability of each study's results to my patient. After reading through the Chen et al and the Deshpande et al articles (Fig. 6: citations 2 and 5), I realized that they did not address my topic specifically enough. Deshpande et al (citation 5) addressed geriatric rehabilitation broadly and Chen et al (citation 2) focused only on subacute rehabilitation, without comparisons to other settings. Therefore, I performed a critical appraisal of articles 1, 3, and 4 listed in Figure 6.

Table. Search History

Search     Search String                                        Results
Line No.

Initial search using keywords (a):
   1       stroke                                                 7,332
   2       acute rehabilitation                                     237
   3       1 and 2                                                   51
   4       from line 3 keep citations 2 (Chen et al),                 4
           13 (Keith et al), 25 (Smithard), 45
           (Schmidt et al) (b)
   5       skilled nursing facility                                 168
   6       1 and 5                                                    5

Revised search using subject headings (c);
   7       acute rehabilitation (with subheadings                16,513
           "rehabilitation centers," "cerebral vascular
           accident," "rehabilitation," or
           functional status)
   8       skilled nursing facility (all subheadings)               907
   9       7 and 8                                                   93
   10      from line 9 keep 18 (Chen et all, 37 (Angelelli            9
           et al), 43 (Kramer et al [2000]), 51 (Skinner),
           55 (Berg and Intrator), 65 (Despande et al),
           72 (Kramer et al [1997, JAMA]),
           76 (Kramer et al [1997, Top Stroke Rehabil]),
           84 (Frolich and Folgelman) (d)

(a) Map Terms to Subject Headings box clicked off.

(b) For citations see Figure 1.

(c) Map Terms to Subject Headings box clicked on.

(d) For citations see Figure 5.

Figure 1. Citations selected from initial search--stroke AND
acute rehabilitation (Table: search lines 1-6). Citations that were
selected for further examination are highlighted in red. Asterisk
indicates a citation that was also retrieved by a search of stroke
AND skilled nursing facility.

1. Chen CC, Heinemann AW, Granger CV, Linn RT, Functional
gains and therapy intensity during subacute rehabilitation:
a study of 20 facilities. Arch Phys Med Rehabil.
2002;83:1514-1523.

* 2. Keith RA, Wilson DB, Gutierrez P. Acute and subacute
rehabilitation for stroke: a comparison. Arch Phys Med
Rehabil. 1995;76:495-500.

3. Smithard DC. Management of stroke: acute,
rehabilitation and long-term care. Hosp Med.
2003;64:666-672.

4. Schmidt J, Drew-Cates J, Dombovy M. Severe disability
after stroke: outcome after inpatient rehabilitation.
Neurorehabil Neural Repair. 1999;13:199-203.

Figure 5. Citations of the articles from revised search--acute
rehabilitation (with subheadings "rehabilitation centers," "cerebral
vascular accident," "rehabilitation," or "functional status")
AND skilled nursing facilities (all subheadings) (Table: search
lines 7-10)--that were relevant for further review. Citations
selected for further review are highlighted in red. Asterisk indicates
a citation retrieved by the initial search.

* 1. Chen CC, Heinemann AW, Granger CV, Linn RT. Functional
gains and therapy intensity during subacute rehabilitation:
a study of 20 facilities. Arch Phys Med Rehabil.
2002;83:1514.

2. Angelelli JJ, Wilber KH, Myrtle R. A comparison of
skilled nursing facility rehabilitation treatment and
outcomes under Medicare managed care and Medicare
fee-for-service reimbursement. Gerontologist.
2000;40:646-653.

3. Kramer AM, Kowatsky JC, Lin M, et al. Outcome and
ultilization differences for older persons with stroke in
HMO and fee-for-service systems. J Am Geriatr Soc.
2000;48:726-734.

4. Skinner N. Acute, subacute, postacute, or skilled
nursing facility: where should your patient go? Inside
Case Management. 1998;5(7):5-8.

5. Berg K, Intrator O. Postacute care following stroke or
hip fracture: single services and combinations used by
Medicare beneficiaries (1987-1992). J Aging Health.
1999;11:27-48

6. Deshpande SA, MacNeill SE, Lichtenberg PA, et al.
Functional outcome differences in acute versus subacute
rehabilitation. Topics in Geriatric Rehabilitation.
1998;13(4):30-38.

7. Kramer AM, Steiner JF, Schlenker RE, et al. Outcomes
and costs after hip fracture and stroke: a comparison of
rehabilitation settings. JAMA. 1997;277:396-404.

8. Kramer AM, Schlenker RE, Eilertsen TB, Hrincevich CA.
Stroke rehabilitation in nursing homes. Top Stroke Rehabil.
1997;41:53-63.

9. Frolich WH, Fogelman L. Short-term rehab maximizes
patient potential. J Long Term Care Adm. 1994;22:4-8.

Figure 6. Articles retrieved from CINAHL that were chosen
for further review based on title and abstract. Citations 1 and
2 were retrieved from the first search (Table: search lines 1-6),
citations 3-5 were retrieved from the revised search (Table:
search lines 7-9). Citations selected for critical appraisal are
highlighted in red.

1. Keith R, Wilson D, Gutierrez P. Acute and subacute
rehabilitation for stroke: a comparison. Arch Phys Med
Rehabil. 1995;76:495-500.

2. Chen C, Heinemann A, Granger C, Linn R. Functional
gains and therapy intensity during subacute rehabilitation:
a study of 20 facilities. Arch Phys Med Rehabil.
2002;83:1514-1523.

3. Kramer A, Kowalsky J, Lin M, et al, Outcome and
utilization differences for older persons with stroke in
HMO and fee-for-services systems. J Am Geriatr Soc.
2000;48:726-734.

4. Kramer A, Steiner J, Schlenker R, et al. Outcomes and
costs after hip fracture and stroke: a comparison of rehabilitation
settings. JAMA. 1997;277:396-404.

5. Deshpande SA, MacNeill SE, Lichtenberg PA, et al.
Functional outcome differences in acute versus subacute
geriatric rehabilitation. Top Geriatr Rehabil. 1998;13:30-38.

* Cinahl Information Systems, 1509 Wilson Terrace, Glendale, CA 91206.

([dagger]) Ovid Technologies, 100 River Ridge Dr, Norwood, MA 02062.

Devon Keane-Miller, DPT, was a student in the DPT program at Boston University, Boston, Mass, at the time this article was written.

Terry Ellis, PT, NCS, is Clinical Assistant Professor of Physical Therapy, Boston University.

Linda Fetters, PT, PhD, is Associate Professor of Physical Therapy and Director, Physical Therapy Programs, Boston University.