CCA MI LLMI LCMI Missing Mechanism Bias ASE Bias ASE Bias ASE Bias ASE MCAR       0.16 0.80 0.07  0.52 0.02 0.54 0.03 0.47  MAR(X2)   0.18 0.93 0.07 0.58  0.05 0.68 0.02 0.54 MAR(Y)     0.19 0.93   0.08 0.54 0.01 0.57 0.03 0.49 MAR(Y,X2)  0.17 0.92 0.04 0.54 0.03 0.60 0.01 0.50  MCAR = Missing completely at random MAR(X2) = Missing at random that depends on x2 (eg. age of the person) MAR(Y) = Missing at random that depends on outcome y (eg. HbA1c) MAR(X2,Y) = Missing at random that depends on outcome Y (eg. HbA1c) and X2 (eg. Age) CCA= Complete case analysis MI = Multiple imputation LLMI = Log-linear multiple imputation LCMI = Latent class multiple imputation with three classes

Table 4: Bias in the mean log-odds ratio (Bias=estimated mean – 0.69) and asymptotic standard error (ASE) estimates of a logistic regression model from a simulation study with 50% missing race data (n=200, true log odds ratio=0.69).