Special Feature

Gastric Tonometry—Research Toy or Clinical Tool?

By Charles G. Durbin, Jr., MD, FCCM

Since the early 1980s, scientists have used semi-permeable, saline-filled balloons equilibrated in the stomach to estimate gastric mucosal pCO2, calculate mucosal pH, and identify inadequate intestinal perfusion. It is hypothesized that inadequate intestinal perfusion reflects incomplete resuscitation or inadequate cardiac output. If ischemia persists long enough, release of gut toxins can lead to the systemic inflammatory response syndrome (SIRS) and multiple organ system failure (MOSF). Gastric tonometry has been suggested as a minimally invasive technique to detect and possibly prevent these prominent causes of intensive care unit (ICU) morbidity and mortality.

This brief essay reviews the information that has accumulated on the technique of gastric tonometry during the past decade. Unfortunately, initial enthusiasm for its routine use in the critically ill has not resulted in dramatically improved patient outcomes. Technical issues, patient selection, and timing of clinical interventions can influence the output and interpretation of this monitor. While gastric tonometry has been useful in increasing our understanding of the pathophysiology of MOSF, its inherent complexity suggests that it is not yet ready for routine clinical use.


Gastric tonometry is performed by placing a small, silastic balloon-tipped catheter in the stomach. Correct placement is usually confirmed by manometry, auscultation, catheter distance calculation, or radiography. The balloon is inflated with saline or other solution, which is allowed to equilibrate for 30-90 minutes, and then the solution withdrawn. The pCO2 is freely diffusible through the wall of the balloon and is assumed to reflect the mucosal pCO2. The pCO2 of the sample of fluid is measured in a conventional blood-gas analyzer; bicarbonate is calculated from an arterial blood gas sample, and, using the Henderson-Hasselbach relationship, mucosal pH (pHi) is calculated.

The sample must be analyzed under strict anaerobic conditions. Some ABG analyzer systems produce inconsistent results with saline and should not be used for gastric tonometry. Histamine 2 (H2) blockade seems to improve consistency. Equilibration time must be strictly maintained, and antacids must be avoided. Currently, only intermittent, labor-intensive measurements of pHi can be made. These difficulties in measurement have prevented the technique from widespread adoption into the clinical environment. However, newer, semicontinuous, automated techniques are on the horizon.

Experimental Data on Efficacy

Is pHi helpful in clinical care? Animal studies have confirmed the value of monitoring pHi. Decreased pHi accompanies hypovolemic, septic, and anaphylactic shock in various animal species. After resuscitation, a delay in return of pHi confirms the sensitivity of mesenteric blood flow to shock states and supports the concept of the gut being the source of toxins. Studies in animals have shown that therapy directed at increasing cardiac output can raise pHi to normal levels in mesenteric ischemia models.

In humans, results have been mixed. During abdominal aortic replacement surgery, pHi fell less with hydroxyethyl starch than with crystalloid despite a similar cardiac output. In a study of critically ill patients, infusions of hydroxyethyl starch were superior to infusions of albumin in moderating decreases in pHi. In one study, low-dose dobutamine was superior to red cell transfusions in raising pHi in septic patients. In a study of bolus administration of fluids, no immediate effect was seen on pHi. Several reports have suggested that dobutamine and dopeximine, but not epinephrine or dopamine, improve pHi in critically ill patients.

In summary, changes in pHi in response to clinical interventions are usually in the direction predicted; however, this is not uniformly the case. The rate of change of pHi is delayed for several possible reasons. The measurement itself requires a long equilibration time and more rapid changes in mesenteric perfusion will not be detected as they occur. There may be specific measurement effects of some of the interventions that have little to do with actual bowel perfusion. In general, changes in pHi related to clinical interventions are not fully understood and should not be used to replace clinical judgment.

Clinical Utility

What about outcome studies? There is no doubt that a low pHi predicts a worse outcome in a variety of circumstances. The classic article by Gutierrez and colleagues demonstrated in a group of critically ill patients that a pHi < 7.32 was associated with a higher mortality than those whose pHi was > 7.32. This predictive value of a low pHi has been confirmed by many others. Chang and others showed that a low pHi 24 hours following resuscitation from trauma was the only factor that separated nonsurvivors from those who survived.

Complications following resuscitation, bowel or vascular surgery, or extracorporeal membrane oxygenator (ECMO) use are associated with a low pHi. Complications and mortality are associated with a low pHi following ruptured aortic aneurysm, during septic shock, and in critically ill trauma patients but not following liver transplantation. In some studies pHi was used to predict the likelihood of weaning from prolonged mechanical ventilation in chronic obstructive pulmonary disease (COPD) patients and the likelihood of developing septic complications in surgical ICU patients (but not survival).

The predictive value of pHi is supported in most studies in which it is evaluated. Finding a low pHi adds information that is not generally available from other patient assessment tools and monitors. However, most studies consist of few patients, and differences in methods of measurement and the definition of a low pHi value make direct pooling of these studies difficult.

Can pHi be Used to Improve Outcomes?

Have attempts directed at altering pHi in a favorable way demonstrated improvement in outcome? So far this has not been fully supported with experimental human data. One large prospective study has been reported. Gutierrez et al reported on the use of a pHi-driven protocol to prospectively treat ICU patients. Two hundred sixty patients were studied. Treatment group patients (n = 135) were aggressively managed if their pHi was 7.35 or lower or if it fell by 0.1 pH units at any time during their ICU stay. In addition to optimizing oxygen delivery variables, filling pressures, and blood pressure, they were treated with volume infusions and, if necessary, dobutamine to return pHi to 7.35 or above. Control patients were monitored with gastric tonometry but treatment of patients was left to the responsible clinician. Eighty-five percent of the treatment group received resuscitation directed by pHi at some time during their ICU stay.

In this study, there was no overall difference in mortality between the treatment and control groups. However, in those patients in whom the initial pHi was normal, the treatment protocol resulted in improved survival (58% vs 42%). A second, small study in trauma patients demonstrated shorter length of stay and fewer complications when low pHi was treated as compared to historical controls. There was no overall improvement in mortality.


Gastric tonometry has contributed to a better understanding of critical illness. It may be clinically useful in certain subgroups of critically ill patients. It seems to be helpful in managing patients who have normal mesenteric perfusion at the time monitoring is initiated. A low pHi identifies a sicker group of patients who are at greater risk of death and complications. However, the technical problems and personnel demands make the use of this monitor unsatisfactory for most ICUs at the current time. Newer techniques may improve user friendliness but the value of the measurement needs further study demonstrating improved outcomes.

The accuracy of gastric tonometry is adversely affected by:

a. H2 blockers.

b. antacid administration.

c. NG tube placed on suction.

d. agitation.

e. narcotic administration.