Experience with on-line computing in clinical chemistry.

Abstract

The scientific practice of medicine is making increasing demands upon clinical chemistry departments and it is the experience of these laboratories, both in teaching and non-teaching hospitals, that their work loads double every four to five years (Lathe and Mitchell, 1966). The use of AutoAnalyzers' has markedly increased the capacity of these laboratories for carrying out repetitive types of analytical work (Table I), and at the same time introduced a degree of much needed uniformity into the methods of performing analyses. Developments in automatic chemical equipment may further increase the work capacity of these departments, but already a different set of problems has emerged. These can be summarized under several interrelated headings: (1) specimen collection and unequivocal identification of samples with the correct patient throughout the subsequent procedures; (2) accession procedures in the laboratory; (3) maintenance of reliable standards of analytical performance in large-scale operations; (4) processing of instrumental readings; (5) report preparation and presentation; (6) uses made of laboratory data (a) by the clinicians and (b) by the laboratory; (7) records storage and arrangements for their retrieval for various purposes. This paper will concentrate particularly on experience gained so far with computer-dependent systems designed to help with problems 2 to 5. Blaivas and Mencz (1967, 1968), reporting upon an extension of their earlier system (Blaivas, 1966), described how an IBM 1710 computer had been routinely used for process control, linked on line to as many as 30 AutoAnalyzers performing up to 20 different analyses. The reasons for wishing to develop an alternative to the system in use at King's County Research Laboratories included the substantial cost (capital or hire) of the IBM system, and the desire to incorporate additional features such as records storage and cumulative reporting of laboratory results using the computer. The system recently installed in this laboratory is intended ultimately to help with problem 7 also, by storing records of laboratory work in a way that will allow the issue of