Knowledge of the pathophysiology and management of primary intracerebral haemorrhage (ICH) has increased substantially over recent years, and the article by Jackson and Sudlow in this issue (see p 1244)¹ makes a further important contribution to our understanding of this serious illness. In a meticulously conducted meta-analysis of observational studies with defined localisation of the ICH, they have shown that overall a history of hypertension was about twice as common in deep ICH as in lobar ICH. However, there was considerable heterogeneity across studies, and sensitivity analyses restricted to higher quality studies markedly attenuated the differential effect of this exposure (or outcome). Thus, apart from …