Original Research
Initial Physical Grades and Cognitive Stages After Acute Stroke: Who Receives Comprehensive Rehabilitation Services? Margaret G. Stineman, MD, Barbara E. Bates, MD, MBA, Jibby E. Kurichi, MPH, Pui L. Kwong, MPH, Diane Cowper Ripley, PhD, W. Bruce Vogel, PhD, Dawei Xie, PhD Objectives: To study the degree to which initial physical grades and cognitive stages of independence assessed by physical medicine and rehabilitation (PM&R) staff early after hospitalization for acute stroke relate to the decision to either provide rehabilitation in consultation or admission to a specialized rehabilitation unit (SRU) for comprehensive, high-intensity, multidisciplinary rehabilitation. Design: An observational study. Setting: Early rehabilitation assessment by PM&R staff during patients’ acute hospitalization for stroke in 112 Veterans Affairs facilities. Patients: The sample included 8,783 veterans who were assessed by PM&R staff. Methods: Shortly after hospital admission, functional status was determined according to 7 physical grades and 7 cognitive stages of increasing independence. Patients’ physical grades and cognitive stages ranged at initial PM&R assessment from the lowest and most dependent “I” through intermediate “II, III, IV, V, or VI,” and ended with the highest at total independence “VII.” To assess the statistically independent effects of physical grade and cognitive stage, a multivariable generalized estimating equation was applied to account for within Veterans Affairs facilities correlation and to adjust for demographics, stroke type, comorbidities, clinical events before PM&R assessment, and facility-related factors. Main Outcome Measurements: The decision to admit patients to an SRU for comprehensive rehabilitation. Results: Only 11.2% of those patients assessed after stroke were admitted to an SRU after the acute management phase. After statistical adjustment, patients at the lowest physical grade (I) of independence had a 9-fold increased odds of admission to an SRU compared with those at the highest combined physical grades VI/VII (adjusted odds ratio 9.15, 95% confidence interval 4.31-19.39). In contrast, patients at intermediate cognitive stages of independence were the most likely to be admitted to an SRU. The presence of an SRU within the treating Veterans Affairs facility was strongly related to admission. Conclusions: Patients’ physical grades and cognitive stages assessed early after stroke are strong determinants of referral for comprehensive rehabilitation. PM R 2013;5:1007-1018
INTRODUCTION A recent American Heart Association/American Stroke Association endorsed practice guideline for the management of adult stroke rehabilitation care highlighted early assessment and treatment as critical to optimizing rehabilitation outcomes [1,2]. In recognizing that maximizing function is a primary goal, the guideline emphasized the importance of applying standardized valid functional assessment tools [3,4] in establishing comprehensive treatment plans. The standard Functional Independence Measure (FIM) instrument (Uniform Data System for Medical Rehabilitation, Amherst, NY), with known validity [3], is applied by the physical medicine and rehabilitation (PM&R) staff within the Veterans Health Administration (VHA) and private sector inpatient PM&R 1934-1482/13/$36.00 Printed in U.S.A.
M.G.S. Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, PA; Department of Physical Medicine and Rehabilitation, University of Pennsylvania, 423 Guardian Drive, 904 Blockley Hall, Philadelphia, PA 19104-6021. Address correspondence to: M.G.S.; e-mail:
[email protected] Disclosure related to this publication: grant, AHRQ B.E.B. Physical Medicine and Rehabilitation, Veterans Affairs Medical Center, Albany, NY; Physical Medicine and Rehabilitation, Albany Medical College, Albany, NY Disclosure: nothing to disclose J.E.K. Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, PA Disclosure related to this publication: grant, AHRQ P.L.K. Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, PA Disclosure related to this publication: grant, AHRQ
Author information continued on page 1018. This project was supported by grant number R01HS018540 from the Agency for Healthcare Research and Quality. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality. It is also supported by resources and the use of facilities at the University of Pennsylvania in Philadelphia, Pennsylvania, the Samuel S. Stratton Department of Veterans Affairs Medical Center in Albany, New York, and the North Florida/South Georgia Veterans Health System in Gainesville, Florida. Peer reviewers and all others who control content have no relevant financial relationships to disclose. Submitted for publication August 27, 2012; accepted August 10, 2013.
ª 2013 by the American Academy of Physical Medicine and Rehabilitation Vol. 5, 1007-1018, December 2013 http://dx.doi.org/10.1016/j.pmrj.2013.08.598
1007
1008
Stineman et al
INITIAL PHYSICAL GRADES AND COGNITIVE STAGES AFTER ACUTE STROKE
rehabilitation facilities (IRFs) during patients’ initial assessments of rehabilitation potential [5] and treatment planning. It is important to address linkages across the complex service bundles that occur during acute hospitalizations. Critical lifesaving and neuroprotective procedures are essential during the hyperacute stage, as are the assessment and rehabilitation of residual functional deficits as soon as medical stability is attained. Standardized electronic records maintained across all Veterans Affairs Medical Centers (VAMC) provide opportunities to study the provision of rehabilitation during acute hospitalizations for stroke. The Functional Status Outcomes Database (FSOD) (VHA Office of Information, Washington, DC) as 1 component of that electronic record, records PM&R assessments, and the results of rehabilitation decision making. This record makes it possible to study rehabilitation decision making in ways that would be impossible in the private sector. Decisions about the type and intensity of rehabilitation services provided at the time of hospitalization are essential for patients with acute strokes, many of whom face new-onset disabilities that could strongly affect their quality of life and capacity to return home. PM&R staff, as documented through the FSOD, makes recommendations to the acute care stroke team with regard to the rehabilitation services needed based on a formal early poststroke assessment. Focused on early discharge planning, these recommendations consider patients’ needs and ability to tolerate therapy, and the availability of resources. Goals established in partnership with the patient and family address strategies for achieving sufficient function for hospital discharge. There are 2 alternative types of rehabilitation services selected during the PM&R assessment, referred to as “consultation” and “comprehensive.” The decision, depending on patient status and goals, may be that sufficient rehabilitation can be provided through consultation by the therapy staff while the patient remains on nonrehabilitation bed services within the acute care hospital. Alternatively, the PM&R team may decide that more intensive rehabilitation is necessary, which requires a transfer to a specialized rehabilitation unit (SRU) bed within the VAMC for comprehensive rehabilitation services. In response to the American Heart Association/American Stroke Association endorsed practice guideline for the management of adult stroke rehabilitation care, we explored how patients’ initial physical grades and cognitive stages of independence assessed early after an acute stroke relate to the decision to either continue consultation rehabilitation or to admit patients to an SRU for higher intensity comprehensive rehabilitation services. Our aim was to study patients admitted to Veterans Affairs facilities for the management of acute stroke with respect to early PM&R assessment and the types of rehabilitation services received while in the hospital. The sequence of increasing functional independence expressed by the physical grade and cognitive stage
thresholds reflect established hierarchies of functioning [6,7]. Physical grades and cognitive stages are sensitive to change over time and are predictive of a variety of outcomes after rehabilitation in various inpatient settings [8-11]. We selected initial FIM physical grades and cognitive stages of independence to study rehabilitation decision making because physical grades and cognitive stages define the maximum levels of assistance needed when performing each of 13 physical and 5 cognitive activities, respectively. Consequently, physical grades and cognitive stages assessed early after acute stroke carry important information that can help in hospital discharge planning and in addressing the scope of rehabilitation services needed to facilitate appropriate discharge. We anticipated that many patients with total physical and cognitive limitations (at the lowest physical grades and cognitive stages of independence) might be too frail physically to tolerate intensive levels of rehabilitation or be too cognitively impaired to be able to fully participate in comprehensive care. Furthermore, we expected that most patients at the highest physical grades and cognitive stages of independence would not require comprehensive rehabilitation. Thus, we hypothesized that decisions to admit stroke patients to an SRU for comprehensive rehabilitation would favor those patients at intermediate physical grades and cognitive stages who need rehabilitation because of severe but not total limitations in function.
METHODS The institutional review boards at the University of Pennsylvania in Philadelphia, Pennsylvania, the Samuel S. Stratton VAMC in Albany, New York, and the University of Florida in Gainesville, Florida, approved this study. It further received approval from the research and development committees in Albany and the North Florida/South Georgia Veterans Health System (VHS) in Gainesville.
Description of Databases The study merged data from 7 VHA administrative databases used to track the health status and health care use by veterans. These databases included the Patient Treatment Files (PTF) (main, procedure, bed section) (Veterans Affairs Information Resource Center, Washington, DC), 2 outpatient care files (visit, event) (Veterans Affairs Information Resource Center), the extended care file (Veterans Affairs Information Resource Center), and the FSOD. The PTF main includes demographic information and inpatient diagnostic information for the entire hospitalization. PTF bed sections capture diagnostic information related to specific hospital specialty bed services that patients received while hospitalized. PTF procedure data contain codes and dates for the procedures received during the hospitalization. Outpatient care files include diagnostic information obtained during outpatient
PM&R
visits. The extended care file also contains comorbidity information while patients were in long-term care facilities. The FSOD has collected FIM information within the VA for many years as part of a quality indicator. It was central to this analysis because it includes information about functional status and details the types of rehabilitation services received.
Study Population All veterans with primary diagnoses consistent with stroke treated in VAMCs and assessed by PM&R staff as recorded in the FSOD over a 2-year period were included in this observational study. We first selected all patients discharged from VAMCs with dates between October 1, 2006, and September 30, 2008, with evidence of a stroke. Patients with stroke who were hospitalized for 365 days or longer were not included. In recognizing that patients are sometimes transferred between centers, PTF admissions or discharges from Veterans Affairs facilities within 1 day of each other were appended to obtain information on the entire acute hospitalization. The hospitalization for the acute stroke, which spanned the PTF admission date through the discharge date, defined the acute “index stroke stay.” Acute stroke was identified based on the presence of specific International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnostic codes included in the stroke high-sensitivity algorithm [12] established for the VA. For patients to be included, these codes needed to be present in the primary diagnostic fields of the patients’ PTF main or bed section records that indicated that cerebrovascular disease was the primary cause of hospitalization. The diagnoses and their corresponding ICD-9-CM codes included occlusion, embolism, or stenosis of the cerebral arteries (434.01, 434.11, 434.91), occlusion or stenosis of the precerebral arteries (433.01, 433.11, 433.21, 433.31, 433.81, 433.91), intracerebral hemorrhage (431.xx, 432.xx), subarachnoid hemorrhage (430.xx), and acute, but illdefined cerebral vascular disease (436.xx). Patients with transient cerebral ischemia (435.xx) were only included if there was additional evidence of stroke, that is, codes that indicated hemiplegia, hemiparesis (342.xx), or 1 or more of the above stroke codes present in secondary diagnostic fields. Patients were excluded if they had evidence of a previous stroke on screening when applying the same set of stroke codes to earlier PTF (documenting a previous hospitalization), outpatient care files, extended care files, or FSOD files that spanned 1 year before the index stroke stay admission date. There were 13,341 patients with an acute stroke, of whom, 8900 individuals were formally assessed by the PM&R staff as documented post hoc by an FSOD record and were eligible for the study. Among these veterans, 117 (1.3%) were missing 1 or more admission FIM scores and were not included in further analyses. A total of 8783 cases were included in the analyses of PM&R decision making from 112 Veterans Affairs facilities.
Vol. 5, Iss. 12, 2013
1009
Variables Outcome. The primary outcome was the type of acute rehabilitation services received among those assessed by the PM&R staff during the index stroke stay. We distinguished between 2 broad types of services. The first was treatment through consultation-level rehabilitation. The second was comprehensive-level rehabilitation (admission to a SRU bed service).
Consultation-level rehabilitation. With consultation, the primary reason for hospitalization remains medical management, but functional restoration becomes an important secondary goal. The consultation-level rehabilitation services provided in nonrehabilitation beds (typically acute medical or neurologic beds) range in frequency and intensity from a single visit to multiple visits. Although the same types of multidisciplinary PM&R staff are available to patients, consultation services are less extensive and more variable in frequency than those that occur with comprehensive rehabilitation services. Comprehensive-level rehabilitation. With comprehensivelevel rehabilitation, the primary reason for hospitalization shifts from medical to functional restoration, and patients are transferred to SRUs from their medical or neurologic bed services. SRUs are distinct rehabilitation bed services located in designated areas within VAMCs. SRUs provide comprehensive inpatient rehabilitation similar in intensity to private sector IRFs and dedicated stroke units. Coordinated multidisciplinary teams of clinical staff provide daily high-intensity “comprehensive” rehabilitative care to patients. SRU teams in the VHA may include the following staff: a physician, nurse, physical therapist, occupational therapist, kinesiotherapist, speech and language pathologist, psychologist, and recreational therapist. VA SRU beds must be accredited by the Commission on Accreditation of Rehabilitation Facilities (CARF), which ensures that staff are capable of providing a high-quality, patientcentered, comprehensive, integrated inpatient rehabilitation program [13]. Patient-level Variables. The primary variables of interest were patients’ physical grades and cognitive stages.
Physical grades and cognitive stages of independence. These measures of functional independence were derived through complex analytic profiling procedures described previously [11,14]. They express the physical and cognitive activities that patients are still able to perform with and without assistance, and the maximum amount of help needed for each FIM activity when assistance is required [3]. Patients’ performances on each of the 18 FIM activities were determined by clinicians during early assessment by PM&R and were recorded in the FSOD. Patients’ physical grades and cognitive stages were derived according to patient performance ratings on the 13 FIM physical items
1010
Stineman et al
INITIAL PHYSICAL GRADES AND COGNITIVE STAGES AFTER ACUTE STROKE
and 5 FIM cognitive items, respectively, known to reliably represent these 2 underlying dimensions [6,15]. Physical grade and cognitive stage definitions and discussions of potential relevance to rehabilitation decision making and discharge care planning are included in Tables 1 and 2. Demographic factors and living circumstances before hospitalization. Demographic factors and living circumstances before hospitalization included age (
