Improving peri-operative care using an anaesthesia information management system

This thesis investigates a reminder system and intends to form a proof of the concept that decision support, using patient and situation specific automated reminders, can actually improve patient outcome. The research in this thesis was planned and executed within the scope of a guideline implementation process. Due to the broad applicability and relevance, we decided on the departmental guideline for postoperative nausea and vomiting (PONV) for our research.
This thesis demonstrated that an automated reminder system, clinically relevant and statistically significant, improves patient outcome by modifying clinician behavior to improve guideline adherence. By using a set of four different automated reminders, we effectively doubled guideline adherence and thereby reduced postoperative nausea and vomiting in high-risk patients from 47 to 31% without increasing the amount of anti-emetics used.
In conclusion, decision support using automated reminders can be an effective tool for supporting guideline implementation, modifying physician behavior and improving patient outcome.
To achieve this the data processing should be designed in a way that there is a minimum of false-positive reminders and as little as possible false-negative situations, where a reminder does not appear when indicated. In addition, the timing, content and the means of effectuating the advice given should be optimally designed. However, even with optimal designs, automated reminders cannot overcome every reason for non-adherence and employing decision support does not replace an implementation and education process.
To prevent unintended deterioration of care, any decision support system employed should be evaluated for its effect.