A 'missing not at random'(MNAR) and 'missing at random'(MAR) growth model comparison with a buprenorphine/naloxone clinical trial
Addiction, 2015•Wiley Online Library
Aims To compare three missing data strategies:(i) the latent growth model that assumes the
data are missing at random (MAR) model;(ii) the D iggle–K enward missing not at random
(MNAR) model, where dropout is a function of previous/concurrent urinalysis (UA)
submissions; and (iii) the W u–C arroll MNAR model where dropout is a function of the
growth factors. Design Secondary data analysis of a N ational D rug A buse T reatment C
linical T rials N etwork trial that examined a 7‐day versus 28‐day taper (ie stepwise …
data are missing at random (MAR) model;(ii) the D iggle–K enward missing not at random
(MNAR) model, where dropout is a function of previous/concurrent urinalysis (UA)
submissions; and (iii) the W u–C arroll MNAR model where dropout is a function of the
growth factors. Design Secondary data analysis of a N ational D rug A buse T reatment C
linical T rials N etwork trial that examined a 7‐day versus 28‐day taper (ie stepwise …
Aims
To compare three missing data strategies: (i) the latent growth model that assumes the data are missing at random (MAR) model; (ii) the Diggle–Kenward missing not at random (MNAR) model, where dropout is a function of previous/concurrent urinalysis (UA) submissions; and (iii) the Wu–Carroll MNAR model where dropout is a function of the growth factors.
Design
Secondary data analysis of a National Drug Abuse Treatment Clinical Trials Network trial that examined a 7‐day versus 28‐day taper (i.e. stepwise decrease in buprenorphine/naloxone) on the likelihood of submitting an opioid‐positive UA during treatment.
Setting
11 out‐patient treatment settings in 10 US cities.
Participants
A total of 516 opioid‐dependent participants.
Measurements
Opioid UAs provided across the 4‐week treatment period.
Findings
The MAR model showed a significant effect (B = −0.45, P < 0.05) of trial arm on the opioid‐positive UA slope (i.e. 28‐day taper participants were less likely to submit a positive UA over time) with a small effect size (d = 0.20). The MNAR Diggle–Kenward model demonstrated a significant (B = −0.64, P < 0.01) effect of trial arm on the slope with a large effect size (d = 0.82). The MNAR Wu–Carroll model showed a significant (B = −0.41, P < 0.05) effect of trial arm on the UA slope that was relatively small (d = 0.31).
Conclusions
This performance comparison of three missing data strategies (latent growth model, Diggle–Kenward selection model, Wu–Carrol selection model) on sample data indicates a need for increased use of sensitivity analyses in clinical trial research. Given the potential sensitivity of the trial arm effect to missing data assumptions, it is critical for researchers to consider whether the assumptions associated with each model are defensible.
Wiley Online Library