Tackling the global burden of stroke: the need for large-scale international studies
To be eff ective, global and regional strategies to reduce the incidence and prevalence of stroke require an understanding of the frequency and determinants of stroke in diverse geographical locations and diff erent ethnic groups. In this issue of The Lancet Neurology, Feigin and co-workers and Johnston and co-workers present important and related epidemiological observations on global stroke incidence. Feigin and co-workers report on the divergence in stroke incidence between low to middle income countries (100% increase) and highincome countries (42% reduction) over the past four decades, whereas Johnston and co-workers show that current approaches used to measure and monitor global determinants of stroke do not provide convincing explanations for these regional diff erences. Given the available information, both articles are exhaustive and complete. However, the data are limited, and were obtained in diff erent ways, and during diff erent periods. Therefore, our ability to compare and integrate information across these studies is restricted. These studies highlight the current defi cits in the study of the determinants of global stroke incidence, particularly in low-income and middle-income countries, in which most strokes occur. Three types of clinical or epidemiological study are needed: large international standardised registries that document the characteristics, presentation, management, and outcomes after stroke, to understand the variations in practice patterns and outcomes in diff erent countries; large standardised case–control studies to clarify whether risk factors for stroke (and its subtypes) vary among regions, ethnic groups, or socioeconomic categories; and large surveillance programmes at sentinel sites in diff erent regions worldwide, to track secular changes in risk factors, key treatments, and stroke rates (eg, SinoMONICA). Despite the importance of stroke as a major publichealth problem, our understanding of the global epidemiology of stroke is incomplete and lags behind our knowledge of conditions such as coronary heart disease. Throughout the past decade, large international studies have advanced our understanding of regional variations in the incidence and risk factors for acute coronary syndromes (eg, INTERHEART) and their management and outcomes (eg, the GRACE and CREATE registries). Such studies have enabled us to tailor populationbased prevention and treatment strategies to such regional variation. Similar studies are now needed in best designs for clinical trials. Additionally, as Collinge and co-workers also note, earlier and improved diagnosis is a crucial advance that will add more power to future trials.