DATA-COLLECTION IN ANESTHESIA - CONCEPT A ND REALITY AFTER 10 YEARS EXPERIENCE

Citation
Am. Zbinden et al., DATA-COLLECTION IN ANESTHESIA - CONCEPT A ND REALITY AFTER 10 YEARS EXPERIENCE, Anasthesist, 46(6), 1997, pp. 496-503
Citations number
12
Categorie Soggetti
Anesthesiology
Journal title
ISSN journal
00032417
Volume
46
Issue
6
Year of publication
1997
Pages
496 - 503
Database
ISI
SICI code
0003-2417(1997)46:6<496:DIA-CA>2.0.ZU;2-N
Abstract
In many institutions information systems are used to process off-line anaesthesia data for invoices, statistical purposes, and quality assur ance. Information systems are also increasingly being used to improve process control in order to reduce costs. Most of today's systems were created when information technology and working processes in anaesthe sia were very different from those in use today. Thus, many institutio ns must now replace their computer systems but are probably not aware of how complex this change will be. Modern information systems mostly use client-server architecture and relational data bases. Substituting an old system with a new one is frequently a greater task than design ing a system from scratch. This article gives the conclusions drawn fr om the experience obtained when a large departmental computer system i s redesigned in an university hospital. Methods: The new system was ba sed on a client-server architecture and was developed by an external c ompany without preceding conceptual analysis. Modules for patient, ana esthesia, surgical, and pain-service data were included. Data were ana lysed using a separate statistical package (RS/1 from Bolt Beranek), t aking advantage of its powerful precompiled procedures. Results: Devel opment and introduction of the new system took much more time and effo rt than expected despite the use of modern software tools. Introductio n of the new program required intensive user training despite the choi ce of modern graphic screen layouts. Automatic data-reading systems co uld not be used, as too many faults occurred and the effort for the us er was too high. However, after the initial problems were solved the s ystem turned out to be a powerful tool for quality control (both proce ss and outcome quality), billing, and scheduling. The statistical anal ysis of the data resulted in meaningful and relevant conclusions. Conc lusions: Before creating a new information system, the working process es have to be analysed and, if possible, made more efficient; a detail ed programme specification must then be made. A servicing and maintena nce contract should be drawn up before the order is given to a company . Time periods of equal duration have to be scheduled for defining, wr iting, testing, and introducing the program. Modern client-server syst ems with relational data bases are by no means simpler to establish an d maintain than previous mainframe systems with hierarchical data base s, and thus, experienced computer specialists need to be close at hand . We recommend collecting data only once for both statistics and quali ty control. To verify data quality, a system of random spot-sampling h as to be established. Despite the large investments needed to build up such a system, we consider it a powerful tool for helping to solve th e difficult daily problems of managing a surgical and anaesthesia unit .