Predictive Analysis

Predictive analytics

Our predictive analytics enable providers to create more personalized healthcare, base their treatments on scientific evidence available and even predict patient behavior. Our insights can also help consumers become more informed about their own health.

  1. Reducing 30-day readmissions
  2. Physician performance management
  3. Hospital Acquired Infections
  4. Evidence Based Guidelines
  5. Predictive LOS
  6. Value Based Purchasing

Reducing 30-day readmissions

Pythia has developed a proprietary statistical platform that can analyze a patient's EMR data in real time and identify those with a high risk of readmission. Case managers receive a complete report with a risk score and the key risk identifiers while the patient is still in the hospital The Report 1 (on admission) would consist of a 30 day all cause readmission risk score and probability for the patient for the index admission The Report 2 would also include susceptibility analysis of the patient for the most common reasons for readmissions, namely CHF, AMI, COPD and PN.

Physician performance management

An evidence-based 360o physician performance assessment including both administrative and clinical data based on a best-in-class statistical model. Our proprietary model normalizes the complexity of procedures and complications of patients to ensure the comparisons are apples to apples Comprehensive appraisal drawn from five performance domains: Process of care, readmissions, productivity, outcomes and patient satisfaction. Our solution is always online and metrics are refreshed monthly to ensure the hospital takes decisions on the latest available information The solution rides on our trusted HIPAA compliant platform ensuring maximum data security. The platform pulls data directly from the hospital EHR via HL7 so there is minimum integration effort and changes to the hospital’s IT systems.

Hospital Acquired Infections

We have developed a scoring system to predict HAI that was derived from Logistic Regression (LR) and validated by Artificial Neural Networks (ANN) simultaneously. This allows infection prevention specialists to efficiently identify patients at high risk for HAI during hospitalization Hospital-acquired infections (HAI) are associated with increased attributable morbidity, mortality, prolonged hospitalization, and economic costs so a reliable prediction model for HAI has great clinical relevance.

Evidence Based Guidelines

We provide an advanced platform for analyzing hospital clinical processes by measuring their adherence to hospital protocols. The system works by collecting and processing data at various decision points of the treatment flowchart. We provide a real-time comparison across specialties as well as across hospital units under same group. All reports are backed up by specialized research evidence and cohort analysis