Persistent gastric intramucosal ischemia in patients with sepsis followingresuscitation from shock

Study objectives: (1) To determine the effects of resuscitation of patients ,vith severe sepsis to conventional hemodynamic end points and normal blood lactate levels on postresuscitation sequential assessments of gastric intr amucosal pH (pHi). (2) To determine whether trends in pHi are reflected in trends in systemic hemodynamic, oxygen utilization, and acid-base assessmen ts. Design: Prospective cohort study. Setting: Medical ICU in an inner-city, university-based medical center. Patients: Twelve recently admitted patients with severe sepsis and signs of circulatory shock who were successfully resuscitated to normal hemodynamic end points and lactate levels and who were also monitored with pulmonary a rtery catheters and gastric tonometers, Interventions: Because of the obser vational nature of this study no specific interventions were employed. The physician staff administered IV fluids and pharmacologic agents, during and after the resuscitative period, to treat infection and to achieve and main tain hemodynamic stability. Mechanical ventilation and supplemental oxygen were provided as needed. The hemodynamic and physiologic monitoring employe d was determined by the managing physicians and established medical ICU rou tines. Measurements and results: A total of 12 patients were studied. Systemic hem odynamic, oxygen utilization, and acid-base assessments and pill were recor ded following resuscitation, and every 12 h thereafter. pHi decreased from 7.33 +/- 0.08 (mean +/- SD) following resuscitation to 7.26 +/- 0.04 at 24 h, 7.20 +/- 0.07 at 36 h (p < 0.05), and 7.24 +/- 0.08 at 48 h. Correspondi ng statistically significant and clinically relevant changes in systemic he modynamic, oxygen utilization, and acid-base variables were not observed. T he hospital mortality of this patient group was high (10 of 12; 83%). Conclusions: Gastric intramucosal acidosis develops and persists for at lea st 48 h in patients resuscitated from septic shock to conventional resuscit ative end points, including the normalization of lactate levels. These regi onal changes were not reflected in corresponding changes in systemic acid-b ase and oxygen utilization variables. Direct determinations of pHi and ther apy directed to toward the resolution of splanchnic ischemia may be require d to improve the outcome in these patients.