Although the diagnosis of bowel ischemia may be done by a variety of different methods, gastric tonometry is the simplest, most practical, and least invasive [6]. It attempts to determine the perfusion status of the gastric mucosa by measuring the local Pco₂ [7]_. Carbon dioxide diffuses from the mucosa into the lumen of the stomach and subsequently into the silicone balloon of the tonometer. The co2 in the balloon is used as a reflection of the mucosal co₂ and can be measured by either saline tonometry or air tonometry. In the saline tonometry technique, a saline solution is injected into the balloon and after an equilibration period, the co₂ is measured using a blood gas analyzer. For the air tonometry technique, air is pumped through the balloon and an infrared detector measures the Pco₂ continuously. As perfusion to the stomach decreases, the Pco₂ in the tonometer will increase. Once cellular anaerobic respiration starts, the hydrogen ions titrate with bicarbonate, with the end result of more co₂ production by mass action. Initially, most investigators calculated the intramucosal pH (pHi) using the Henderson-Hasselbach equation and assumed that arterial bicarbonate was equal to the mucosal bicarbonate. Experimental evidence has shown that a lower pHi was associated with lower mucosal flow by laser Doppler, demonstrating that mucosal bicarbonate is not equal to arterial bicarbonate [8]. Furthermore, it has been shown that respiratory acid/base disturbances affect pHi calculation [9]. Therefore, the use of the Pco₂ gap (the difference between gastric mucosa and arterial co₂) is considered a better way to assess gastric perfusion [10].

The use of the technique has not gained widespread popularity despite many clinical studies that have validated gastric tonometry as a valuable and easily accessible prognostic tool [11, 12]. This may be explained by the possibility of error in the determination of the Pco₂ and interoperator variability [13, 14]. Other pitfalls include multiple local effects, including increased gastric secretions and refluxed duodenal contents; both of which can increase co₂ measurement and lead to false Pco₂ measurement, and that this technique may only represent one region of perfusion [7].

Laser Doppler flowmetry, which estimates gastric blood perfusion by integrating red blood cell content and velocity, is highly correlated with absolute blood flow. The flowmeter consists of a laser source, a fiberoptic probe, and a photodetector with a signal-processing unit. The laser conducts through the tissue by a flexible fiberoptic guide. The probe contains an optic fiber for transmission of laser light to the tissue and two fibers for collecting the reflected scattered light. The signal-processing unit consists of a photodetector and an analog circuit to analyze the frequency spectrum of the scattered light. By determining the instantaneous mean Doppler frequency and the fraction of backscattered light that is Doppler shifted, the signal-processing unit provides a continuous output proportional to the number of red blood cells moving in the measuring volume and the mean velocity of these cells. Measurements are considered satisfactory if: (a) the measurement is stable for 15 seconds; (b) the measurement is free of motion artifacts; (c) pulse waves can be clearly identified; and (d) the reading is reproducible [8].

Near infrared spectrometry (NIRS) has been used to measure local tissue blood flow and oxygenation at the cellular level [15]. Local oxygen delivery and oxygen saturation can be determined by comparing the differences in absorption spectrum of oxyhemoglobin with its deoxygenated counterpart, deoxyhemoglobin [16]. Puyana et al. [17] reported using NIRS to measure tissue pH in a model of experimental shock and showed that NIRS gut pH correlated with the pH obtained by microelectrodes.

Positron emission tomography (PET) may also be used to evaluate regional blood flow. Fluoromisonidazole accumulation has been used to demonstrate abdominal splanchnic perfusion and
regional oxygenation of the liver in pigs; however, the lack of portability of this technique makes it difficult to use for monitoring in the intensive care unit (ICU) [15].

Microdialysis measurement of mucosal lactate is a novel way to assess gut mucosal ischemia. Tenhunen et al. [18] inserted microdialysis catheters into the lumen of the jejunum, the jejunal wall, and the mesenteric artery and vein of pigs. Subsequently, the animals were subjected to nonischemic hyperlactatemia or an episode of mesenteric ischemia and reperfusion. The lactate levels from the jejunal wall and the jejunal lumen were compared. The gut wall lactate was increased in both the nonischemic and the ischemic lactatemia, whereas the lactate measured from the jejunal lumen only was altered significantly during true ischemia.

Microdialysates of other substances have also been measured, including glucose and glycerol, showing that, while lactate levels increase with ischemia, intestinal wall glucose levels drop with the same stressor. Glycerol was increased, but the changes were seen later than the changes in lactate [19]. Similarly, increases in the lactate/pyruvate ratio in both intraperitoneal or intraluminal placed microdialysis catheters have correlated with hypoperfusion [20]. As glucose from the splanchnic circulation is inhibited, pyruvate accumulates in the tissue and, in the setting of inadequate oxygen delivery, is broken down to lactate. Using glycerol as a marker, Sollingard et al. [21] suggested that gut luminal microdialysis could serve as a valuable tool for surveillance not only during ischemia, but also after the ischemic insult.

These data support the idea that microdialysis could be a potentially useful method to monitor gut ischemia. However, even under investigational conditions, technical difficulties were reported in up to 15% of cases by either damage to the microdialysate membrane, dislocation of the probe, or incorrect placement [22].

The D-xylose test has been used in the diagnosis of malabsorption. This pentose sugar of vegetable origin is incompletely absorbed in the small intestine by a passive mechanism. The test consists in the ingestion of 25 g dose of D-xylose and the subsequent measurement of the levels on the serum or in urine. In normal individuals, a serum sample taken 1 to 2 hours after ingestion will reveal a level of 25mg per dL and a 5-hour urine collection will result in at least 4 g of this substance. Many entities such celiac disease, alterations in gastrointestinal motility, and impaired function of the pylorus will result in abnormal results. In the critically ill, renal function may be altered and may alter the results of the urine test. Chiolero et al. [26] studied the intestinal absorption of D-xylose in the critically ill patients that were tolerating enteral feeding. They introduced D-xylose to the stomach or the jejunum and found that although the levels in plasma in all patients in the study increased, indicating proper gastric emptying, in those receiving the compound in the stomach, the levels of D-xylose were lower than normal, indicating delays or depression in absorption. These results were similar to a prior study in trauma and septic patients. In this study, in both groups the D-xylose test showed abnormal results at the onset of the illness with resolution by 1 to 3 weeks after trauma or resolution of sepsis. Interestingly, enteral feedings were tolerated by these patients before the test results returned to normal [27]. As the patients in both studies were tolerating tube feeds even with abnormal D-xylose test results, Chiolero et al. [26] suggested that this test may not be a good indicator to determine the capacity of patients to tolerate enteral feeds. This does confirm that absorption of D-xylose stays depressed for a prolonged period of time in the critically ill.

Johnson et al. [28] also found decreased absorption in the septic population when compared with healthy individuals. They used an oral test solution that contained 5 g of lactulose, 1 g of L-rhamnose, 0.5 g of D-xylose, and 0.2 g of 3-O-methyl-D-glucose. L-rhamnose is absorbed by passive diffusion and therefore particularly sensitive to changes of the absorptive capacity of the gut when compared with D-xylose and 3-O-methyl-D-glucose, which depend on specific carrier mechanisms. The authors found that septic patients had decreased L-rhamnose/3-O-methyl-D-glucose ratios when compared with normal individuals, a result consistent with decrease absorptive capacity during sepsis. They also used the lactulose/L-rhamnose ratio to assess permeability of the gut. This group concluded that the
changes in the absorptive capacities of the gut may contribute to the pathophysiology of sepsis.

The rapid absorption of acetaminophen at the jejunal level can also aid the assessment of the absorptive capacity of the gut. It has, however, been used more to assess gastric emptying [5] and tube feeding location for enteral feeding [29]. Details of this test were discussed in the motility section. From these data, it appears that either carbohydrate absorption tests or the acetaminophen test could be used to a monitor absorption in the critically ill. No correlation has been established between tolerating tube feeds and the degree of absorption. The role of this test may be to monitor improvement of absorptive function of the gastrointestinal tract after critical illness.