Development and validation of prognostic and diagnostic prediction models

Document Type

News Article


Objective: Development and Validation of Multivariate Prognostic Model based on Risk Factors and Inflammatory Markers to Predict Outcome in Neonatal Sepsis Patients.

Methodology: A prospective observational cohort study, conducted in Kasturba hospital, Manipal, Karnataka by taking a cohort group i.e., neonatal sepsis. The patient will be followed till the discharge and at the end will be divided into two group i.e., patient with and without mortality. Prognostic prediction model will be developed separately for early onset and late onset sepsis by using patient related factors (age, sex, gender etc) and biochemical parameters (CRP, IL6). Prognostic prediction model will be developed based on logistic regression.

We hypothesize that the prediction model developed by logistic regression will be useful in identifying various risk factors which significantly contribute towards neonatal mortality in sepsis.

  1. Prediction modelling in the early detection of neonatal sepsis- https://link.springer.com/article/10.1007/s12519-021-00505-1#:~:text=Prediction%20modelling%20approaches%20were%20able,in%20the%20years%20to%20come.
  2. Identification of potential risk factors for the poor prognosis of neonatal sepsis https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9387580/
  3. Identification of risk factors associated with hyponatremia in psychiatric patients: A case control study. https://medpharmareports.com/index.php/mpr/article/view/2368

Publication Date

Spring 10-30-2022