Abstract
It is important to base the practice of physical therapy on evidence accumulated from clinical trials. To achieve this, physical therapists need an “information system” consisting of two aspects: 1) a large scale database of clinical trials, and 2) a good search engine to allow effective access to the database. To explore the current feasibility of such a system, we used 3 types of data sources: (i) the Cochrane review database, (ii) a textbook written from this premise (Clinical Decision Making in Rehabilitation), and (iii) raw data selected from Medline searches. From these, we retrieved the literature reporting the evidence of physical therapy and arranged them according to the categories of disease or disorder, therapeutic means, outcomes, effectiveness, and by the level of evidence. Fifty-nine of a total of 1, 854 abstracts registered in the Cochrane Library related to physical therapy or rehabilitation. In 25 reports, the benefit of treatment was shown statistically significant. Although the level of evidence was not high in several studies, 41 trials contained in the book have demonstrated the effectiveness of treatment. Using Medline, we searched for literature from 1997 to 2000 and found 37 as relevant. In 31 of them the effectiveness of the treatment was shown at statistically significant levels. During Medline searches, we also recognized the importance of accurate keyword selection to obtain proper literature for our purpose. Knowledge of where and how to find evidence is sparse, and we cannot effectively utilize whatever evidence is already stored in the databases. Therefore, we must not only accumulate evidence to support physical therapy, but we have to construct the database and search-engines for access, as well. If such a database were to be realized, we could then compare different types of physical therapy and select the optimal treatment regimens.
