Meta-analysis is a method of synthesizing the results of independent s
tudies. We consider the case in which there are multiple treatments an
d a control, with the goal of estimating the relative effect of each t
reatment based on continuous outcomes. Even when all data are availabl
e, rather than only summary data, it has become common to use meta-ana
lytic estimators of treatment contrasts. Alternatively, we could use a
two-way analysis of variance model with no interaction in which one f
actor is study and one factor is treatment. For the unbalanced case, w
e obtain the surprising result that the standard meta-analysis estimat
es of treatment contrasts are identical to the least squares estimator
s of treatment contrasts in the linear model. Because a meta-analysis
of individual patient data can be considerably more costly in terms of
data retrieval than a meta-analysis of summary data, this equivalence
provides for cost-efficient analysis.