Structure, process, and outcomes in stroke rehabilitation.

BACKGROUND: The health services research framework of structure, process, and outcome is used commonly to examine quality of care, and it indicates that structure influences process, which in turn influences outcomes. However, little empirical work has been done to test this hypothesis, particularly for medical rehabilitation. OBJECTIVES: To determine if, among stroke patients, (1) structure of care was associated with process of care, and (2) structure of care was associated with outcomes after adjusting for process. RESEARCH DESIGN: Two-year, prospective study of 288 acute stroke patients in 11 VA medical centers, of whom 128 were included in the current analysis. MEASURES: Structure of care: systemic organization, staffing expertise, and technological sophistication. Process of care: compliance with the AHCPR poststroke rehabilitation guidelines. PATIENT CHARACTERISTICS: baseline prior walking ability and Functional Independence Measure (FIM) motor subscale. OUTCOMES: the FIM motor subscale 6-months poststroke. RESULTS: The combination of systemic organization and staffing expertise, along with technological sophistication, were independent predictors of process of care (beta coefficients 0.21, P<0.05 and 0.37, P<0.001, respectively). When controlling simultaneously for patient characteristics, structure and process of care, structure of care did not have and process of care did have a statistically significant association (beta coefficient 0.18, P<0.01) with functional outcomes. CONCLUSIONS: Better process of care was associated with better 6-month functional outcomes, therefore improving process of care probably would improve stroke outcomes. However, our results indicate that improving key structure of care elements might facilitate improving process of care for stroke patients.

Postacute stroke guideline compliance is associated with greater patient satisfaction.

OBJECTIVE: To determine if the structure of care or the process of stroke care, as measured by compliance with stroke guidelines published by the Agency for Healthcare Research and Quality (AHRQ), is associated with patient satisfaction. DESIGN: Prospective inception cohort study of new stroke admissions including postacute care with follow-up interviews at 6 months poststroke. SETTING: Eleven Veterans Affairs medical centers (VAMCs). PARTICIPANTS: A total of 288 new stroke patients admitted to VAMCs. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Compliance with AHRQ stroke guidelines and patient satisfaction with care using a stroke-specific instrument. RESULTS: Process of care was positively and significantly associated with greater patient satisfaction even after controlling for patient functional outcome. The most visible (to the patient) process of care dimensions correlated most highly with patient satisfaction. Sixty-four percent (73/115) of patients expressed some dissatisfaction with 1 or more survey items. CONCLUSIONS: "What we do" and "how we do it" while providing postacute care to stroke patients was associated with patient satisfaction. This linkage of process to outcome is an important validation of satisfaction as a significant patient outcome. This linkage is further evidence that compliance with AHRQ stroke guidelines may be a valid quality of care indicator.

Measurement characteristics of the Functional Independence Measure.

Clinicians and researchers recognize the need for measures offunctional status that possess linear properties and are reliable and valid. Rasch rating scale analysis provides the means for converting raw scores from functional assessment tools to linear measures for which measurement error can be quantified. The extent to which clinicians perceive patients who are undergoing rehabilitation after stroke as similarto other patient groups was investigated using the Functional Independence Measure (FIM). Earlier work demonstrated that the first 13 items of the FIM represent a measure of motor function and that the last 5 items represent a measu re of cognitive function. The FIM was used for patients with stroke in a manner similar to that for most other impairment groups on the motor items. Patients with stroke were, however, unlike many impairment groups in their ratings on the cognitive items. Tables showing raw score to scaled measure conversions are provided for two sets of impairment groups on the motor items and three sets of impairment groups on the cognitive items. Clinicians can be confident that the measures derived from the FIM are linear across the range of the instrument and are attuned to the uniqueness of patients with stroke and other specific impairments.