Allowing for uncertainty due to missing data in meta‐analysis—Part 2: Hierarchical models

We propose a hierarchical model for the analysis of data from several randomized trials where some outcomes are missing. The degree of departure from a missing-at-random assumption in each arm of each trial is expressed by an informative missing odds ratio (IMOR). We require a realistic prior for the IMORs, including an assessment of the prior correlation between IMORs in different arms and in different trials. The model is fitted by Monte Carlo Markov Chain techniques. By applying the method in three different data sets, we show that it is possible to appropriately capture the extra uncertainty due to missing data, and we discuss in what circumstances it is possible to learn about the IMOR.

[1]  S. Sharp,et al.  Explaining heterogeneity in meta-analysis: a comparison of methods. , 1999 .

[2]  K. Barry,et al.  Brief physician advice for alcohol problems in older adults: a randomized community-based trial. , 1999, The Journal of family practice.

[3]  Gareth O. Roberts,et al.  Non-centred parameterisations for hierarchical models and data augmentation. , 2003 .

[4]  P. Gustafson.  On Model Expansion, Model Contraction, Identifiability and Prior Information: Two Illustrative Scenarios Involving Mismeasured Variables , 2005 .

[5]  P. Anderson,et al.  The effect of general practitioners' advice to heavy drinking men. , 1992, British journal of addiction.

[6]  J. Ockene,et al.  Brief physician- and nurse practitioner-delivered counseling for high-risk drinkers: does it work? , 1999, Archives of internal medicine.

[7]  D J Spiegelhalter,et al.  Bayesian approaches to random-effects meta-analysis: a comparative study. , 1995, Statistics in medicine.

[8]  Elizabeth A. Stasny,et al.  Hierarchical Models for the Probabilities of a Survey Classification and Nonresponse: An Example from the National Crime Survey , 1991 .

[9]  E Scott,et al.  Randomized controlled trial of general practitioner intervention in women with excessive alcohol consumption. , 1991, Drug and alcohol review.

[10]  A. Haines,et al.  Randomised controlled trial of general practitioner intervention in patients with excessive alcohol consumption. , 1988, BMJ.

[11]  Roderick J. A. Little,et al.  Modeling the Drop-Out Mechanism in Repeated-Measures Studies , 1995 .

[12]  Alex J. Sutton,et al.  Multiparameter evidence synthesis in epidemiology and medical decision‐making: current approaches , 2006 .

[13]  Douglas G Altman,et al.  Statistical methods for assessing the influence of study characteristics on treatment effects in ‘meta‐epidemiological’ research , 2002, Statistics in medicine.

[14]  R. Little.  Pattern-Mixture Models for Multivariate Incomplete Data , 1993 .

[15]  D. Altman,et al.  Measuring inconsistency in meta-analyses , 2003, BMJ : British Medical Journal.

[16]  Daniel O Scharfstein,et al.  Incorporating prior beliefs about selection bias into the analysis of randomized trials with missing outcomes. , 2003, Biostatistics.

[17]  A Wodak,et al.  Controlled evaluation of a general practice-based brief intervention for excessive drinking. , 1995, Addiction.

[18]  Alexander J Sutton,et al.  What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. , 2004, Statistics in medicine.

[19]  Sander Greenland,et al.  Multiple‐bias modelling for analysis of observational data , 2005 .

[20]  Carrol Gamble,et al.  Uncertainty method improved on best-worst case analysis in a binary meta-analysis. , 2005, Journal of clinical epidemiology.

[21]  Roderick J. A. Little,et al.  Statistical Analysis with Missing Data: Little/Statistical Analysis with Missing Data , 2002 .

[22]  Stephen Rollnick,et al.  Lassa fever: epidemiology, clinical features, and social consequences , 2004, BMJ : British Medical Journal.

[23]  Balgobin Nandram,et al.  A Bayesian analysis of a proportion under non‐ignorable non‐response , 2002, Statistics in medicine.

[24]  A. Gelman.  Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper) , 2004 .

[25]  K. Barry,et al.  Brief physician advice for problem alcohol drinkers. A randomized controlled trial in community-based primary care practices. , 1997, JAMA.