Inpatient quality assurance by local analysis of uniform data set data

We constructed a data base in order to examine the profile of patients admitted to a 24 bed inpatient rehabilitation unit at Fairfield Hospital. Data were collected according to the Uniform Data Set for Medical Rehabilitation, and collated and analysed using the Clinical Reporting System database software and the Statistical Package for Interactive Data Analysis statistical software. Outcome data collected included the patients domicile before and after admission, the duration of stay, medical diagnoses before admission to and during the stay in the Rehabilitation Unit, as well as the Functional Independence Measure item scores at admission and discharge. Analysis of the first 100 patients in the data base indicated a duration of stay of 43 +/- 34 days with a right skew distribution. Scatter plots of age and duration of stay, and age and functional improvement showed no bias in the Unit concerning these parameters. The Wilcoxon rank sum test indicated a highly significant difference (P = 0) between entry and exit Functional Independence Measure scores, indicating that patients improved functionally in the unit. Eighty three patients returned home on discharge and only five required admission to a nursing home. With the data base, the course of patients whose performance lies outside the range of normal for the unit can be examined and factors identified which produce a prolonged stay or failure to progress.