Transporting databanks of medical information from one location to another

If ever clinical information systems, particularly those used to support decision-making, are to become of more than local value, there is an urgent need to develop 'transportable' databanks of information - that is to say, databanks which can be utilised for one purpose or another at different locations geographically without degradation of their effectiveness. This presentation discusses the problems caused by the need to develop 'transportable' databanks. First, the rationale for their need and for their use is discussed. It is pointed out that increasing mobility of patients and doctors demands increasing standardisation and harmonisation of clinical terminology and databanks of information. Next, the present situation is reviewed in some detail. The current lack of 'transportable' databanks is described, and the effects of this set out. Specifically, the factors which prevent current databanks being used outside their own centre of origin are listed; such as differences of hardware and software, differences in perceived and actual needs, inadequate sampling, analysis and testing, and geographical variation in disease spectra and disease presentation. The consequence of this need is a need to develop and test multi-national databanks of information, and this presentation next defines what is needed in order to produce such databanks. Areas discussed include the definition of the clinical problem, construction of proformata and collection of data, matrix construction and testing, and finally wide distribution. In conclusion, the presentation defines some of the future perceived needs in respect of databanks and some of the problems which will arise in their construction. It is concluded that action is necessary now on a multinational basis in order to develop the databanks of information - truly transportable - which will have become mandatory by the 1990's.