Traditionally, shock has been recognized or diagnosed by subjective signs and symptoms, particularly in septic shock, where transition from localized to systemic infection and then to septic shock may be gradual and subtle. Management has been directed toward normalizing these subjective symptoms as well as BP, heart rate, urine output, hematocrit, central venous pressure, and blood gases. The major problem is that restoration to normal values of these secondary aspects of shock do not correct the underlying tissue perfusion defect. The aim of this review is to describe a physiologic mechanistic model based on the concept that uneven vasoconstriction and maldistribution of flow are directly related to tissue hypoxia, oxygen debt, shock, shock-related organ failure, and death; second, to show that titration of therapy to optimal physiologic end-points using hemodynamic and oxygen transport monitoring is a potentially cost-effective therapeutic approach. This physiologic approach is based on the hypotheses that: a) the physiologic patterns of high-risk postoperative and septic survivors are significantly different from septic nonsurvivors; b) tissue perfusion can be evaluated by the sequential patterns of cardiac index, oxygen delivery (DO2), and oxygen consumption (VO2) measurements; c) the observed increased cardiac index and DO2 in the survivors are compensations that improve tissue oxygenation, which is reflected by the VO2 pattern; and d) the supranormal values that were documented in survivors provide objective physiologic criteria for therapeutic goals. The data suggest that a mechanistic analysis of the pathogenesis of shock may be elucidated by temporal patterns of the nonsurvivors' physiologic variables. That is, the predictive indices calculated for each variable quantitatively reflect the relationship of the early changes leading to death or survival. In essence, early changes in those variables statistically related to death may reflect pathogenic mechanisms, while early changes related to survival may be used as a first approximation to therapeutic goals. The application of this approach in prospective, randomized trials has demonstrated that prompt attainment of optimal goals (empirically defined from survivors' patterns) improved outcome in postoperative shock with and without sepsis, as well as in medical sepsis and accidental trauma. Specifically, when the optimal values of cardiac index, DO2, and VO2 used as therapeutic goals were attained in 8 to 12 hrs, there was marked and significant reduction in mortality and morbidity rates. This finding was also confirmed in 12 prospective, controlled trials, four of which were randomized. We conclude that driving septic shock patients into the survivors' patterns improves outcome, as has been shown in other shock syndromes.(ABSTRACT TRUNCATED AT 400 WORDS)