AI Clinical Pathways/Decisions

World-leading AI clinical pathways/decision technologies.

Clinical Performance Measurement and Standardization within units and across sites enhancing patient safety and clinical outcomes.


Deontics is an Artificial Intelligence (AI) company with world-leading clinical pathway/workflow and clinical decision support technologies. Deontics technologies are the results of long-term academic research in medical AI originating at Carnegie Mellon University, and subsequently developed at the Medical Research Council (UK), Cancer Research UK, Oxford University and University College London. Deontics technologies were developed from first principles for clinical use incorporating concepts from cognitive psychology and computer science. Deontics can utilize descriptive and predictive analytics outputs to create prescriptive analytics solutions. By integrating individual patient data from the EMR with guideline/protocol knowledge encoded in the ProForma clinical logic language, Deontics pathway and decision solutions render relevant, personalized evidence items extracted from selected clinical guidelines and local protocols at a highly granular level at the point of care in the relevant workflow context and increase compliance with clinical guidelines in practice. Whenever any input data items change, the system will instantly (nanoseconds) reevaluate the relevant evidence and patient data and render new outputs. Deontics solutions, by promoting personalized evidence based medicine, provide the capability to implement performance standardization and increase patient safety both at unit level and across sites or regions effectively combating unwarranted variabilities in clinical practice. Deontics solutions will simultaneously integrate individual patient data with local clinical quality standards and publish alerts where a specific patient's treatment or data indicate that they are in breach of such standards. All clinical data items can be functionally linked with standard ontology codes e.g. SNOMED, RxNorm, LOINC etc., which allows for use of ontological logic within decision functionality as well as curation, communication and interoperability benefits. Similarly clinical procedures could be linked with CPT and ICD9 and ICD10 or equivalent codes introducing potential financial parameters into the record, and potentially the decision logic (according to client requirements). Clinical test data can be run against Deontics pathways enabling high volume batch mode analytics on pathways and clinical systems including patient risk-stratification. Because decision data (including rationales and guideline compliance status) are captured and timestamped alongside outcomes and all other clinical data i.e. the 'why' as well as the 'what', Deontics provides the capability to fully ‘measure’ clinical performance at both an individual and population level and provide benchmarks for improvement. The Deontics technology suite includes sophisticated authoring tools which convert selected paper guidelines and local protocols into computer readable format and create workflows without writing computer code which are then integrated with individual patient data. Because there is no necessity to write computer code, it is a very quick process to create a new implementable clinical pathway or to update an existing pathway. Because no computer skills are necessary to create or update new pathways, client clinicians can easily create their own pathways. All data items flowing through Deontics systems are captured and time-stamped and are available for audit and research purposes and can be analyzed individually or at population level in real-time.  Deontics has developed specific tools for multidisciplinary teams and tumor boards. Deontics can also enable patients to have access to relevant personalized evidence and to take part in shared decision making. Deontics systems have been implemented in major US and UK hospitals.

Register to view this innovation

  • Access all of the innovation information
  • Contact the innovator
  • Follow this innovation's progress
  • View innovation updates