Why They Are Different: Based on the Burden of Disease Research of WHO and Institute for Health Metrics and Evaluation
Table 1
Update of DisMod-MR.
DisMod-2
DisMod-MR
DisMod MR 2.0
(i) Parametric assumption with single number (ii) Needing more flexible model for estimation (iii) Adjusted covariate variables (iv) Lack of meta-analysis, Bayesian inference, regression
(i) Bayesian analysis (a) Flexible model (b) Markov Chain Monte Carlo for hierarchical model (c) Availability of uncertainty interval (ii) Meta-analysis: global, regional fit (iii) Regression: adjusted age, comorbidity, etc. (iv) UI of “p X RR” (v) Adjusting the local fit using global fit with cascade