The efficiency of computerised clinical decision support systems

• Published in: The Lancet (British edition) 2024-02, Vol.403 (10425), p.410-411
• Main Authors: Amaral, Andre Carlos Kajdacsy-Balla, Cuthbertson, Brian H
• Format: Article
• Language: English
• Subjects: Artificial intelligence; Clinical decision making; Decision support systems; Decision Support Systems, Clinical; Drug dosages; Drug interaction; Drugs; Efficiency; Heart; Heart rate; Hospitals; Humans; Intensive care; Intensive care units; Large language models; Patients; Quetiapine; Torsades de pointes
• ISSN: 0140-6736; 1474-547X; 1474-547X
• DOI: 10.1016/S0140-6736(23)02839-8

In one study, clinicians had to review 123 alerts to prevent a single adverse event.3 Several solutions exist for decreasing alert fatigue, such as streamlining the list of drug interactions, learning from past overridden alerts, and focusing such alerts on medications less commonly used.4 In The Lancet, Tinka Bakker and colleagues5 used a stepped-wedge cluster trial to test the effectiveness of a streamlined CDSS to decrease the number of high-risk drug combinations in nine intensive care units (ICUs) in the Netherlands. [...]alerts would only be brought to the clinician's attention if the integration of all these factors was clinically important or if the continuously monitored heart-rate corrected QT interval increased beyond a set limit where the risk of Torsades de Pointes was high.12 Furthermore, it would not simply alert the clinician that the patient is taking QT-prolonging drugs, but would present data on the known effects on the heart-rate corrected QT interval of each drug and suggest alternatives that might be less likely to prolong it (for example, quetiapine is unlikely to increase the heart-rate corrected QT interval).13 With the addition of large language models, the system would also have read the clinical knowledge to understand why a specific drug is being used and automatically override alerts for drugs that are either essential or provide the clinician with a better alternative and rationale for them. [...]recommendations from a CDSS to tailor antimicrobial therapies were accepted in only 38% of patients in an ICU because of the need for a more integrated evaluation that was undertaken by the clinician.14 It is necessary to foster and fund future work using the robust methods from the study by Bakker and colleagues,5 but using more advanced and intelligent CDSSs.