Rehabilitative placement of poststroke patients: Reliability of the Clinical Practice Guideline of the Agency for Health Care Policy and Research

Objective: To determine the interrater reliability of the United States Age ncy for Health Care Policy and Research (AHCPR) Clinical Practice Guideline Number 16 for rehabilitative placement of poststroke patients. Design: Pairs of rehabilitation professionals, highly trained in the Guidel ine, rated the appropriateness of rehabilitative placements. Setting: Acute care hospitals in three regions of the country. Patients: Sixty patients with moderate-to-severe stroke. Measures: Numerous factors affecting appropriate placement according to the Guideline were abstracted from medical records or obtained by direct evalu ation of patients. Results: Good reliability was attained for home and nursing facility placem ent with rehabilitation services but with no multidisciplinary rehabilitati on program (intraclass correlation coefficient =.73 and .60, respectively). Serious reliability problems were found for placements in low-intensity ou tpatient rehabilitation and high-intensity inpatient rehabilitation program s. Chief sources of unreliability were ambiguous or missing data in hospita l medical records, complexities in the Guideline. and raters' tendencies to follow their own clinical judgments. More than one type of placement was a ppropriate for 65% of patients. Conclusions: Reliable placement guidelines are possible, but aspects of the Guideline require additional development. Evidence of demonstrated reliabi lity and validity will be required to resolve disputes between rehabilitati on professionals and payers regarding appropriate levels and types of rehab ilitation and to guide patients and their families.