TheImplementation ofDataWarehouse andOLAPforRehabilitation OutcomeEvaluation: TheReDWinESystem

Wecreated adatawarehouse andOLAPsystem on thewebforoutcome evaluation ofrehabilitation services. Thirteen outcome indicators wereuseinthis research. Efficiency oftherapists andclinics, expected utility oftreatments andgraphic patterns weregeneratedfor data exploration, data mining and decision support. Users canretrieve plenty ofgraphs andstatistical tables without knowing database structure orattributes. Ourexperiences showed that multi-dimensional database andOLAPcould serve asadecision support system. Center forRehabilitation Service andtheDepartment of Physical Therapy, Schoolof Healthand Rehabilitation Science, University ofPittsburgh, performed anoutcome evaluation ofrehabilitation services around Pittsburgh area. Thirteen outcome indicators, including Activities inDaily Living Scale, Oswestry LowBackDisability Questionnaire, SF-36, etc., wereusedintheresearch. Thegoals included identifying thefactors predicting outcomes of patients, comparing theeffectiveness oftreatments and monitoring theperformance of specific institutions andtherapists. Material andMethods A casemanagement system wasimplemented on 1997andprovided clinic data ofpatients. Evaluation formsweretransmitted toresearch center and indicators wereinput intoanother database. We chose Microsoft SQLserver TM and OLAP serverTM tobuildthemulti-dimensional database. Transact-SQLTM wasusedtocreate starschemaandmergemanagement/evaluation data.