Köth H., MD, Backhaus C., Friesdorf W. MD
Institute for Human Factors, Technical University Berlin, Steinplatz 1, D-10623 Berlin
Objective. Hospitals find themselves under budget constraints and are forced to optimise their workflow and to redesign the medical pathways. In order to support medical processes, new hardware is often introduced, such as new infusion pumps or modern ventilators.
With new functions implemented, a more effective treatment and workflow is supposed to save time and money. In this context „total intravenous anaesthesia” has become popular. A new approach combines infusion pumps to a functional unit, which is controlled by a monitor to adjust each single pump. The advantage of such a unit is in a huge amount of functions for every attached device to support the needs of a TIVA.
Methods. To evaluate medical devices, the single assessment of functionality is insufficient even though it’s the only criteria in order to select a certain product.
To support clinical workflows complex medical devices need to have a combination of functionality and easy handling. But often high functionality impedes the access even to obvious functions. Thus a real benefit for clinical work is possible by user orientated engineering.
We investigated a total intravenous anaesthesia unit by a participative analysis during the work of 12 physicians. The anaesthetists were held to perform a narcosis under standardised circumstances with two relevant complications to deal with an anaesthesia simulator. After the anaesthetists finished the narcosis, they were told to fill out a standardised questionnaire.
By analysing the workflow, visualisation and verification by experts, we could define the key points and required functions of a TIVA. Afterwards, we compared these requirements with the functions available and analysed the problems the users detected to find a certain function and to execute them.
Results. An interaction with the single pump directly is complicated by a (main) central control. Most common under time pressure this causes a major time delay. E. g. the application of a bolus infusion needs several tasks instead of just pressing one or two buttons at the pump. On the other hand some functions were rated very useful like the takeover of data and the possibility to set rules for a TIVA from induction to reversal.
Conclusion. The problems detected while using the system mostly resulted from difficulties to find a function within many possible selections. Minor and major important functions are not separated. By accentuation of user oriented functions the usability could be improved. The set up of a feedback loop from user to systems engineer would help to differentiate.