Trauma Resuscitation Errors and Computer-Assisted Decision Support

Abstract

Computer-aided, real-time decision support resulted in improved protocol compliance and reduced errors and morbidity. Trial Registration clinicaltrials.gov Identifier: NCT00164034.

Assessing trauma care provider judgement in the prediction of need for life-saving interventions

“…Within this general debate, the clearest role for math-based systems development-computer modelling and instrumentation based on this work-is in improved data-gathering and decision-assist tools [27,28]. Our findings support the notion that a specific high-impact area for such research and decision-support development is in prehospital transport care, trauma centre patient reception, and resuscitation [2]. The chief limitations in this study are those selection biases classically inherent in any kind of voluntary, conveniencesampling-based, human-subjects research.…”

“…Tools comprising computer models and other automated predictors could provide much needed objective data to improve both PPV and agreement across providers. The prediction errors and provider agreement observed in this study may indicate that quality improvement measures and decision-assist tools could reduce errors of LSI commission and omission found in trauma patient resuscitation at major trauma centres and could enhance prompt decision-making capabilities [2] in austere and forwarddeployable settings. Further study on the impact of human judgement on outcome is warranted, as well as investigation into tools to augment and improve prediction of life saving interventions.…”

“…Since it is well known that even in trauma centres with considerable expertise, decision-support reduces errors of omission and commission during the first 30 min of reception and resuscitation [2]; decision-support tools could perhaps be useful to coordinate the team activities. Although pre-hospital care providers had the most time with each patient through the processes of extraction and transport, they were within the most constrained data-gathering circumstances before being asked to make their prognoses.…”

“…However, as research and systems development progresses towards the increasing automation of monitoring and decisionassist technology [2][3][4][5], particularly to support trauma care in forward-deployed, field, and austere settings, the factoring-in of clinical judgement constants has lagged. Human judgement is complex, influenced by a myriad of social and individual factors, and very difficult to quantify in a manner accessible to math-based systems.…”

Considerations for online deviation detection in medical processes

“…For example, Fitzgerald et al designed and deployed a process guidance system in a trauma center to guide medical professionals during the first 30 minutes of trauma resuscitation [13]. This system increased compliance with the underlying medical algorithms and reduced error rate, but these algorithms did not support complex process behaviors, such as concurrency and exception handling.…”