Statistical Methods in Research with Mortality Outcome in Longitudinal Study

Abstract

Data on mortality are useful for medicine and public health. A part of
data is from research. The statistical methods that suggested for analysis of
mortality data in longitudinal study are Poisson and survival analysis. However,
using statistical methods have not been appropriated.
This systematic review aimed to describe the use of statistical methods
for analysis of mortality data in the longitudinal studies. The articles published
in Thai medical and public health journals ranked in the top 50 Thai journal
impact factor during 20132015 were reviewed. The articles with mortality
outcome and longitudinal design were included. The statistical methods
used in articles were evaluated. Frequency and percentage were used to
analyse.
Fourteen articles met the inclusion criteria. Ten articles used survival
analysis and four did not use. Two articles used frequency and percentage;
one used logistic regression, and one used ttest with chisquare test.
In conclusion, most of Thai medical and public health articles used
survival analysis for analyzing mortality data in the longitudinal studies which
was appropriate method and according to recommendation.