@neuRisk demonstrator version 2.5

This prototype implements a collection of prototype decision support and other clinical services based on the new Disease Management Model developed in WP4.3. This is a comprehensive and formal model for capturing and storing demographic and other patient information which is relevant to risk assessment and treatment It consists of a workflow model, a data model, a logic model and a knowledge database. More information can be found in the documentation uploaded under WP4.3 at the @neurIST collaborative space.

Decision support module

@neuRISK prototype version 2 decision support module was built around the risk factors that are present in the existing literature. Deliverable 18 reviewed systematically the existing systematic reviews and highlighted 22 risk factors related with the rupture of an unruptured cerebral aneurysm and subarachnoid haemorage in general.

The results presented by D18 are mostly in the form of Relative Risk, Odds Ratios and Hazard Ratios. They provide useful insight for the disease, but they are not suitable for a quantitative risk-benefit analysis that is required to produce accurate evaluation of the risks or benefits of a treatment. Therefore in this prototype we followed a quantitative, rule (or argument) based approach to present them.

How the risk factors were interpreted

The size (RF5) and the location (RF4) of the aneurysm were treated as the most important risk factors because they are the only ones that we have absolute risk probabilities coming from ISUIA study. For size and location combinations with low risk the module suggests surveillance and for combinations with high risk, it suggests treatment. When size and location combination do not provide a clear benefit for treatment, the rest of the risk factors become important.

Risk factors 1, 2, 3, 6, 7 (Sex, Age, Ethnicity, Presence of symptoms, Previous SAH) provide relative risk about the risk of rupture of an unruptured aneurysm and they were suitable for affecting the decision.

Risk factors 8 to 22 (except 12 - Polcystic kidney disease) are about the risk of SAH in the population just having the factor and without knowledge of bearing an aneurysm or not. Without further epidemiological analysis they are not suitable to use in the decision module. However since this is a prototype showing how knowledge can or could be used, we encoded these risk factors but with minimum impact to the outcome of the decision.

Detailed comments about the D18 analysis and applicability of the results will be circulated in an internal delivarable available in the collaborative workspace.