Demonstrations of query development with Hydra.

Using GUI the graphical user interface.

Terminology and notes

The registry displayed in these videos contains a large number of data services. Each service has its own function, accessing often disparate sources of data to return a result i.e. ‘retrieve all instances of drug administration for a specific patient’. Together these data services constitute a virtual database which is constantly refreshed each time the query is run. The model clinical database alluded to in these examples is made up of anonymised data constituted from real databases at a US tertiary medical facility in order to test the technology and so is a realistic representation of how Hydra would work.

This short video demonstrates how our query tool, Hydra can be used to formulate a query that may serve as a diagnostic variable for example, for sepsis monitoring. In this demo the diagnostic variable will be a test for recent administration of heparin based drugs. This may be useful for example, if a patient exhibits cardiopathy and the clinical team wish to ensure that this is not simply because the patient has been given heparin based drugs recently. Cardiopathy can then serve as a diagnostic variable for sepsis. Again, combining diagnostic variables like this one, with others, allows explanations for the patient’s presentation to be systematically ruled out. Hydra is capable of supporting as many as 28 of these diagnostic variables.

This short video demonstrates how our query tool, Hydra can be used to formulate a federated query across several data sources including a European therapeutic classification of drugs called the Anatomical Therapeutic Chemical (ATC) classification system and a Canadian database of drug products called Health Canada Drug Product Database (DPD). The question we want to answer is: What selective immunosuppressants are available in Canada? This type of query may be useful to inform research or clinical trials or to explore new drug treatment options once first and second line options have been exhausted. This type of query also demonstrates Hydra’s flexibility in being able to federate data from a combination of public and restricted data sources.

This short video demonstrates how our query tool, Hydra can be used to formulate a query that may serve as a diagnostic variable for example, for sepsis monitoring. In this demo the diagnostic variable will be a test for a pre-existing diagnosis of Diabetes. This may be useful for example, if a patient presents with a high blood sugar level and the clinical team wish to ensure it is not simply because the patient is suffering from Diabetes.

While an individual diagnostic variable is useful, by combining diagnostic variables like this one (and those that follow in other demos), explanations for the patient’s symptoms are systematically ruled out. Hydra thereby offers exponentially growing power, in terms of sensitivity and specificity, to automatically detect possible cases of sepsis.

This short video demonstrates how our query tool, Hydra can be used to formulate a query that may serve as a diagnostic variable, for example for sepsis monitoring. In this demo the diagnostic variable will be the presence of hyperglycemia, which is one of the sepsis related symptoms provided the patient does not have Diabetes. As before, by combining diagnostic variables like this one, other explanations for the patient’s presentation are systematically ruled out and Hydra offers exponentially growing power, in terms of sensitivity and specificity, to detect possible cases of sepsis. IPSNP can link the query results to dashboards and flags on EPRs in order to provide automated decision support.

This short video demonstrates how our query tool, Hydra can be used to formulate a federated query across several data sources including the US Government website on clinical trials (ClinicalTrials.gov). In this demo we want to answer the following question. We are interested in a bacteria, salmonella enterica serovar typhi and we want to know: What drugs have been tried in treatment of infections caused by this bacteria (i.e. typhoid fever)? Again, it is possible to glimpse the power of Hydra to support the development or treatment options or to keep pace with new drug discovery and thereby offer non-knowledge base decision support tools. Again it is possible to see how reinforcing query variables increase exponentially the power of Hydra to return answers to complex questions.