model based health change monitoring in pre-surgical patients
DESCRIPTION
Model based health change monitoring in pre-surgical patients. Jan 21, 2014. Petros Endale May 18, 1985 B.Sc in computer science(2006) M.Sc Telemedicine and e-health(2014 ) Primary Advisor Prof Gunar H. Background. 12,000 annual elective surgery at UNN - PowerPoint PPT PresentationTRANSCRIPT
Model based health change monitoring in pre-surgical
patients
Jan 21, 2014
Petros Endale
May 18, 1985
B.Sc in computer science(2006)
M.Sc Telemedicine and e-health(2014)
Primary Advisor Prof Gunar H.
Background• 12,000 annual elective surgery at UNN• Population settlement of Northern Norway• The overall risk of surgery is low in healthy
individuals. Preoperative tests usually lead to false-positive results, unnecessary costs, and a potential delay of surgery. Preoperative tests should not be performed unless there is a clear clinical indication.
Surgery cancellation
Patient68% (8309)
Hospital non-clinical
24% (2980)
Hospital clinical8% (986)
2001-2002(a Hospital in UK)
e-Team Surgery…MSc Project
• Exploring if moving the pre-surgical planning out of hospitals and to patients at home through electronic collaboration will improve the quality of care for patients scheduled for surgery
• Developing system for monitoring health changes in pre-surgical patients. The focus will be on the patient model
Patients
GPs
SurgeonsAnesthesiologist
How can we identify serious changes in the patients health remotely?
Challenges• Due to the vast scope of pre-operative
assessment, the clinical domain knowledge potentially relevant for assessment is virtually limitless• a comprehensive list of co morbidities, full
history of previous surgery, medication, family history, allergies, previous experiences of clinical adverse events
• Data availability
Preop
Detect deviations• Questionnaire(baseline data) + Objective
physiological parameters• Rule Engine(based on guidelines and
expert opinion)• Notification, Recommendation and status • The Health professional decision and
action• The patient status aproved by the HP or in
agreement with the previous model is taken as the new model.
Patient status
Risk scores
Self Assessment
GuidelinesRules
Models
Inference Engine
Status
Models• Model can be seen as a simplified high-
level description of a specific patient in XML form.
• The model can inform patients and physicians about the status of the patient, and deviation from expected/normal
XMLRule
Database
Patient
HealthNet
Server
• Questionnaire answers and physiological data coming from the patient
• The rule engine and the reasoner compare it with previous models and determines the current state of the patient
• The current state along with the decision of the health professional will be saved as the new model
• Anonimze and save the model the model for future similar cases(case based reasoning)
Future work• If significant number of Models and their
related decisions are collected then automatic statically population model can be developed.