Statistical Methods for Cardiovascular Researchers

Abstract

Rationale: Biostatistics continues to play an essential role in contemporary cardiovascular investigations, but successful implementation of biostatistical methods can be complex.

Objective: To present the rationale behind statistical applications and to review useful tools for cardiology research.

Methods and results: Prospective declaration of the research question, clear methodology, and study execution that adheres to the protocol together and serve as the critical foundation of a research endeavor. Both parametric and distribution-free measures of central tendency and dispersion are presented. T testing, ANOVA, and regression analyses are reviewed. Survival analysis, logistic regression, and interim monitoring are also discussed. Finally, common weaknesses in statistical analyses are considered.

Conclusions: Biostatistics can be productively applied to cardiovascular research if investigators (1) develop and rely on a well-written protocol and analysis plan, (2) consult with a biostatistician when necessary, and (3) write results clearly, differentiating confirmatory from exploratory findings.

Keywords: biostatistics; regression analysis; research; research design; survival analysis.

Figure 1

Example of a box plot.

Figure 2

Flowchart of distribution free analysis for continuous data

Figure 3

Flowchart of regression analyses.

Figure 4

Flowchart of parametric analyses for analysis of variance.