Strict blood glucose (BG) control is proved to improve the outcome in patients with glucose intolerance both in acute and
chronic phases, irrespective of whether the patient has diabetes mellitus. However, strict BG control by conventional methods
is so complicated that it cannot be performed easily in normal clinical situations. Furthermore, it is sometimes inadequate.
Therefore, a clinically applicable, reliable artificial pancreas (AP) has long been sought after for more than 40 years. Considering
the present important situations concerning AP, a survey of recent progress in AP is highly desirable. In this review, recent
progress in mechanical AP (MAP) and in MAP-related items is presented. MAP is composed of three major components: a BG control
algorithm, a drug administration system, and a glucose sensor. Recent progress in development of these components is presented,
followed by descriptions of representative MAPs. Although significant progress in the development of MAP has been made, its
use in clinical situations is limited or for research purposes at present. The main limiting factor is the slow progress in
the development of glucose sensors. However, more widespread clinical application of the MAP will occur in the near future,
considering the number of reliable long-life intravenous glucose sensors under development. Another factor is the worldwide
recognition of the importance of BG control in acutely ill patients, in whom the period of strict BG control is usually for
several days to a few weeks.
Key words Artificial pancreas - Blood glucose control - Insulin therapy - Diabetes mellitus - Critically ill patients
This review is revised from the article written by the author in the Japanese Journal of Artificial Organs 2006;35:341–345