Table of Contents:
  • Paper-based patient records
  • Computer-based patient records
  • Aims of a patient data management process
  • Data, information, and knowledge
  • Single tables and their limitations
  • Multiple tables: where to put the data, relationships among tables, and creating a database
  • Relational database management systems: normalization (Codd's rules)
  • From data model to database software
  • Integrity: anticipating and preventing data accuracy problems
  • Queries, forms, and reports
  • Programming for greater software control
  • Turning ideas into a useful tool: eNICU, point of care database software for the NICU
  • Making eNICU serve your own needs
  • Single versus multiple users
  • Backup and recovery: assuring your data persists
  • Security: controlling access and protecting patient confidentiality
  • Asking questions of a data set: crafting a conceptual framework and testable hypothesis
  • Stata: a software tool to analyze data and produce graphical displays
  • Preparing to analyze data
  • Variable types
  • Measurement values vary: describing their distribution and summarizing them quantitatively
  • Data from all versus some: populations and samples
  • Estimating population parameters: confidence intervals
  • Comparing two sample means and testing a hypothesis
  • Type I and type II error in a hypothesis test, power, and sample size
  • Comparing proportions: introduction to rates and odds
  • Stratifying the analysis of dichotomous outcomes: confounders and effect modifiers; the Mantel-Haenszel method
  • Ways to measure and compare the frequency of outcomes, and standardization to compare rates
  • Comparing the means of more than two samples
  • Assuming little about the data: nonparametric methods of hypothesis testing
  • Correlation: measuring the relationship between two continuous variables
  • Predicting continuous outcomes: univariate and multivariate linear regression
  • Predicting dichotomous outcomes: logistic regression, and receiver operating characteristic
  • Predicting outcomes over time: survival analysis
  • Choosing variables and hypotheses: practical considerations.