This repo lists resources for understanding and building risk prediction models in healthcare, especially involving time-to-event data, also called survival analysis. There is a huge volume of scientific literature, and books published on this topic. The intention is to list those that are accessible. To a certain extent, I shall also list resources that includes modeling data involving repeated measurements (longitudinal analysis).

• Survival Analysis: A Self-Learning Text - a succintly written book in simple English, for someone who wishes to understand survival analysis conceptually
• Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis - read this book after having a conceptual understanding
• Medical Risk Prediction Models - this book will teach you to ask the right questions, and prepare you for the analysis of prognostic models
• An Introduction to Statistical Learning - Chapter 11 deals with survival analysis
• Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating

The following are the widely used R and Python libraries.

• Statistical methods for risk prediction and prognostic models
• Survival analysis
• AI for medicine specialization

• Calculating the sample size required for developing a clinical prediction model
• A Predictive Model for Progression of Chronic Kidney Disease to Kidney Failure - this paper applies survival analysis technique to understand the prognosis of chronic kidney disease. It uses electronic health records (EHRs) of patients with kidney disease to predict their disease progression. The creator of this model has exposed their API freely through web browser - the famous Kidney Failure Risk Equation (KFRE). The model parameters can be obtained through their online supplement
• External validation of a Cox prognostic model: principles and methods - this paper describes the steps required to externally validate a Cox prognostic model
• Lessons learnt when accounting for competing events in the external validation of time-to-event prognostic models - this paper discusses the importance of accounting for competing risks when validating the progonostic models externally. This paper has also provided their R codes explaing how to validate a competing-risk model
• The harm of class imbalance corrections for risk prediction models: illustration and simulation using logistic regression
• Assessing performance and clinical usefulness in prediction models with survival outcomes: practical guidance for Cox proportional hazards models
• Prognosis of chronic kidney disease with normal-range proteinuria: The CKD-ROUTE study
• Classifier technology and the illusion of progress
• Random Survival Forests - Random forest method applied to time-to-event data. Provides a detailed description of the algorithm
• WTTE-RNN: Weibull Time To Event Recurrent Neural Network - this paper introduces Recurrent Neural Network (RNN) technique to model time-to-event outcomes in the case of discrete or continuous censored data. There is a Keras implementation of this technique here
• Machine learning for ECG diagnosis and risk stratification of occlusion myocardial infarction

• Survival analysis in R - a good introductory tutorial on survival analysis using R
• External validation of a Cox prognostic model - this article describes (with R code) external validation of Cox regression model as described in this paper
• Frank Harrell's blog - emphasizes the importance of statistical thinking and posts about regression modeling strategies, clinical trials, drug evaluation, medical diagnosis, and decision making.
• Bayesian survival analysis for "Game of Thrones"

• Frank Harrell
• Richard Riley
• Karandeep Singh
• Maarten van Smeden
• Navdeep Tangri
• Allen Downey
• Rami Krispin
• Cam Davidson-Pilon