Skip to main content

Table 5 MCAR and efficacy rate 98 % versus 95 % (RD 0.030): estimated efficacy differences, coverage and bias for 5 %, 15 % and 30 % averages of number of simulated data sets that converged of 5000 data sets, 50 imputations

From: Is using multiple imputation better than complete case analysis for estimating a prevalence (risk) difference in randomized controlled trials when binary outcome observations are missing?

Model

No. of data sets*

RD (RMSE)

Coverage

Bias

All outcomes recorded:

5,000

0.030 (0.026)

0.939

0.000

5 % of outcomes missing

    

CC

5,000

0.030 (0.026)

0.940

0.000

MI: wt, hb, age, para

4,977

0.028 (0.027)

0.948

-0.002

MI: hb, age, para

4,982

0.028 (0.027)

0.957

-0.002

MI: hb, age, para, group

4,992

0.030 (0.027)

0.953

0.000

MI: wt, hb, age, para, group

4,988

0.030 (0.027)

0.951

0.000

15 % of outcomes missing

    

CC

5,000

0.030 (0.028)

0.944

0.000

MI: wt, hb, age, para

4,962

0.026 (0.029)

0.970

-0.004

MI: hb, age, para

4,969

0.025 (0.029)

0.970

-0.005

MI: hb, age, para, group

4,984

0.031 (0.030)

0.957

+0.001

MI: wt, hb, age, para, group

4,972

0.031 (0.030)

0.961

+0.001

30 % of outcomes missing

    

CC

5,000

0.030 (0.030)

0.940

0.000

MI: wt, hb, age, para

4,896

0.021 (0.033)

0.969

-0.009

MI: hb, age, para

4,937

0.021 (0.033)

0.978

-0.009

MI: hb, age, para, group

4,948

0.031 (0.037)

0.971

+0.001

MI: wt, hb, age, para, group

4933

0.031 (0.038)

0.967

+0.001

  1. *Number of data sets for which convergent analysis was achieved