The principle of pulse oximetry is based on two light sources with different wavelengths (660 nm and 940 nm) emitted through the cutaneous vascular bed of a finger or earlobe. The Hb absorbs more light at 660 nm and HbO2 absorbs more light at 940 nm. A detector at the far side measures the intensity of the transmitted light at each wavelength and the arterial oxygen saturation (SpO₂) is derived by the ratio between the red light (660 nm) and the infrared light (940 nm) absorbed[7]. Erikoglu et al and DeNobile et al measured oxygen saturation of the bowel by pulse oximetry and investigated the relationship between these measures and concomitant pathological grading. They found that intraoperative evaluation of intestinal viability by pulse oximetry may give an idea about the degree of pathological changes[8,9]. MacDonald and coworkers supported the notion that a pulse oximeter has the potential to be of value in the intraoperative assessment of intestinal blood flow[10]. A reflectance pulse oximetry of the colon mucosa was applied in an animal model during aortic reconstruction. It was shown to be useful in monitoring the blood flow of the distal colon[11]. Pulse oximetry has been shown to be applicable as an adjunct technique in the assessment of small bowel ischemia in a patient with a strangulated ileus[3]. In a recent study, a novel reflectance pulse oximetry technique was introduced for intraoperative measurements of SpO₂ in the esophagus and large bowel. The method has been proven to be able to measure SpO₂ continuously in patients with a compromised peripheral perfusion[12]. However, according to Hadley and Mars, pulse oximeters do not measure blood flow or tissue viability; they only measure hemoglobin saturation. A low oximeter reading in the limbs would usually be interpreted as a reflection of inadequate central oxygenation and the authors doubt that the pulse oximetry readings taken from the bowel are different from that taken from the limbs. They agree that pulse oximetry could be, at best, an adjunct to established techniques[13]. Dyess et al[14] argued against the clinical value of pulse oximetry given it was found to result in a high rate of false-negative and false-positive evaluations.

The principle of tissue oxygen tension (PtO₂) measurements relies on the reduction of molecular oxygen at a noble metal cathode affecting the current set-up in a polarized circuit. The current that is set up by the probe is linearly dependent on the partial pressure of oxygen in the tissue being measured. The PtO₂ is an extravascular parameter as oxygen in the tissue is measured after capillary gas exchange. It correlated well with StO₂ and IVM findings under stable respiratory conditions[15]. Sheridan et al measured PtO₂ on the colon of 50 patients undergoing colonic resection and anastomosis and found the decreasing PtO₂ levels at the colonic serosa to be predictive of anastomotic leakage[16]. Conversely, in a study by Jacobi et al, no decrease of the submucosal PtO₂ was seen in patients with anastomotic insufficiency after esophageal resections. Even increased PtO₂ values were observed in the anastomotic leakage group[17]. These controversial findings have led to doubts about the role of impaired tissue oxygenation in anastomotic healing.

Spectrophotometry (spectroscopy) uses the principles of light transmission and absorption to measure the concentrations of Hb oxygen saturation in tissues (StO₂). The amount of light passing through the substance is dependent upon the wavelength; thus, near-infrared and visible light have different characteristics for spectrophotometry. Visible light spectrophotometry (VLS) relies on locally absorbed, shallow-penetrating visible light at 475-625 nm wavelengths and the near-infrared is the part of the spectrum with wavelengths just above the visible, typically beginning at 700-730 nm wavelengths. Visual light penetrates approximately 2 mm into the tissue, whereas near-infrared light has a deeper penetration. As a result, near-infrared spectrophotometry (NIRS) provides a global assessment of oxygenation in all vascular compartments (arterial, venous and capillary) and VLS is designed for monitoring of StO₂ in the capillaries[7,18]. In combination with an intravenous dye injection, NIRS can be used to directly measure perfusion [19]. Visible light spectroscopy is similar to NIRS in that the mean VLS StO₂ is in accordance with NIRS StO₂ (bias VLS–NIRS -1 ± 5%; P = NS)[20].

The spectroscope or oximeter emits low-powered white or near-infrared light from a handheld or endoscopic probe placed near or on the bowel wall. The light penetrates and diffuses and reemerges colored according to the oxygenation level (StO₂). Tissue Hb is estimated as [deoxyhemoglobin + oxyhemoglobin], and the tissue Hb oxygen saturation (StO₂) is determined as [oxyhemoglobin]/[deoxyhemoglobin + oxyhemoglobin][20]. VLS and NIRS have been applied in recent studies.

Hirano and colleagues measured bowel StO₂ at the serosa using NIRS in a pilot study[21]. StO₂ was measured at the anastomotic site of the bowel in 20 patients during colorectal resections for colorectal cancer. It was shown that StO₂ of the anastomotic site can be safely and reliably measured by NIRS during colorectal surgery. Low StO₂ on both sides of the anastomosis may indicate an increased risk of anastomotic complications, although further study is needed to determine the cutoff value for StO₂ to prevent serious complications.

Lee and colleagues assessed colon mucosal oxygen saturation during aortic reconstruction surgery in a prospective observational study of 25 patients using a VLS oximeter with an endorectal spectrophotometer probe. In their opinion, intraoperative VLS is a sensitive measure and persistently low colon mucosal oxygen saturation suggests colon ischemia and a threat of colon infarction[22].

Karliczek et al evaluated the predictive value of VLS for anastomotic leakage of the colon and the rectum in 77 patients undergoing colorectal resections in a prospective observational study[1]. StO₂ levels in colonic tissue were shown to be stable and reproducible. Rising values of colon serosal StO₂ were observed at the anastomotic site of the bowel after the construction of anastomosis. The mean StO₂ value at the proximal anastomotic end in the group of patients who healed uneventfully was 72.1 ± 9.0% and it increased to 76.7 ± 8.0% after construction of the anastomosis. The authors speculate that the underlying mechanism of increasing tissue oxygenation is a response to ischemic preconditioning caused by manipulating the bowel. There was no rising StO₂ in anastomoses that ultimately leaked. In anastomoses showing leakage, a significantly lower StO₂ was recorded in the cecum (73.6 ± 5.7 in non-leakage vs 69.6 ± 5.6 in leakage group).

Compared to VLS, the minimal sampled volume of the tissue by NIRS is relatively large, whereas in VLS the point measurements of small tissue samples are possible[1,18]. Advocates of VLS claim that a shallow penetrating visible light is more appropriate for the measurements of bowel wall oxygenation[1]. Another advantage of VLS is that, compared with NIRS, the normal range is significantly narrower: ± 3% vs± 9% respectively[20,23]. A tissue contact is not required in VLS measurements as the instrument corrects for an uneven baseline and the full light spectrum is analyzed[1]. This is a great advantage compared to other techniques such as polarographic or tonometry measurements, NIRS or laser Doppler flowmetry (LDF) where the probe opposition to the tissue leads to decreased perfusion and StO₂[20]. Although there are no comparison studies, VLS is a newer technique for bowel oxygenation measurement, applied in the majority of studies and seems to be more promising.

Oximetry has a number of drawbacks. Firstly, a specific level of StO₂ that leads to intestinal tissue ischemia has not been defined to date. In studies where VLS of the skin was applied, the critical StO₂ level of microsurgical flap perfusion was between 10% and 15% and a level of 15% saturation has been shown to be clinically sensitive and specific in determining amputation level[24,25]. Secondly, there is no uniformity in StO₂ measurements as different algorithms are used by different oximetry systems for estimating StO₂. A variety of VLS and NIRS equipment is available where the measurements are performed with different amount of wavelengths and also at different wavelengths of light[7,26,27]. Thirdly, the reproducibility of the results can be affected by the presence of bile, stool or food within the intestine which can interfere with the passage of light when measurements are taken at the mucosa. Also, the optimal angle of the probe relative to the tissue is 90º and a different angle can lead to inaccurate measurements[28]. Measurements are not always feasible for colorectal anastomoses within 5 cm from the anal verge or when the site of measurement is situated behind sponges and retractors[1]. Finally, prices of the oximetry systems are currently high.

Intravital microscopy (IVM) has been considered a gold standard for microcirculatory research because it can directly visualize and quantify changes at the capillary level using fluorescent-labeled plasma or blood cells[15]. Yasumura et al performed microscopy of the serosal bowel layer and calculated the ratio of blood cell transition and the effective area of the vascular bed. These parameters were found to be useful indices for prediction of bowel survival[29]. This technique has not yet been introduced into human studies, probably because it is time consuming.

In earlier studies, intraoperative Doppler ultrasound of the marginal arteries was believed to be a more reliable intraoperative predictor of intestinal viability than clinical assessment alone[30]. The authors favored the low costs and simplicity of the technique and suggested its adjunctive use[31]. In a study of 117 patients undergoing intestinal anastomosis or enterostomy, a Doppler ultrasound was applied to determine the adequacy of blood supply at the margins of resection. No postoperative complications were experienced when relying on a Doppler signal[32]. In a more recent series of colorectal resections in 200 patients, only 1% incidence of anastomotic insufficiency was observed when the bowel ends to be resected were assessed with a Doppler ultrasound[33]. Conversely, other authors observed that this adds little if anything to clinical judgment[34]. Sensitivity of Doppler ultrasound of 86% was shown in an experimental study which was significantly lower than that of laser Doppler and perfusion fluorometry[31]. Doppler ultrasonography resulted in a high rate of false-negative and false-positive results in an experimental study by Dyess et al[14]. The superiority of laser Doppler flowmetry compared with ultrasound studies was shown[35]. Widely discussed limitations of this technique are that it is considered to be vulnerable to the signals from the neighboring large vessels, it requires an artery exposure and a pulsatile blood flow and a tissue contact is required which can impair local blood flow[1,29].

Hydrogen gas clearance was attempted for intestinal microcirculation assessment in earlier experimental studies[36,37,38]. The basic paradigm of hydrogen clearance consists of inserting a positively polarized electrode into tissue, administering H2 either by respiration or intra-arterially, allowing the H2 to be cleared from arterial blood, and then monitoring the exponential clearance rate of H2 from the tissue [39]. This technique has been validated for measuring blood flow in different organs; however, it is not used routinely for evaluation of intestinal ischemia due to its invasiveness, inconvenience and insufficient accurateness[21,29].

Radioisotopes were used in earlier studies to quantify the perfusion at the anastomotic or ischemic site of the bowel. In animal and human studies, intraperitoneal[40], intravenous[41,42], intra arterial[43] or local submucosal[44] injections of radioisotopes were applied to monitor ischemic sites and blood flow in the intestine or anastomosis. Disadvantages of radioactive isotopes are exposure to radiation for both the patient and personnel, the technique is expensive in terms of storage and disposal and it is rather cumbersome[45]. These are the possible causes why this technique is not employed at present.

Two techniques, the perfusion fluorometry and laser fluorescence angiography (LFA), have been applied for evaluation of intestinal viability. Perfusion fluorometry was tried in animal and human studies and gained wide acceptance in the diagnosis of acute bowel ischemia in earlier studies. Here, the sodium fluorescein is administered intravenously and the bowel is illuminated with an ultraviolet light investigating the perfusion. It can be used both for a laparotomy when the source of UV light is a Wood’s lamp as well for laparoscopy when the optical filters are placed to the light source of the laparoscopic set to produce UV light[46-50]. The earlier studies applied qualitative fluorometry and in the more recent studies, a fiberoptic fluorometry technique was introduced to measure the blood flow in dye fluorescence units. Its superiority over the qualitative technique was shown[50,51]. Previous studies reported adverse reactions of intravenous fluorescent dye injections, even anaphylactic reactions with fatal outcomes[52,53]. However, a recent international multicenter study has shown intravenous administration of fluorescein to be safe for a confocal laser endomicroscopy of the gastrointestinal tract[54]. The major drawback of fluorometry is the inability to perform repeated measurements since the fluorescein sticks around once it gets into the tissue[50].Very large standard deviations of quantitative fluorometry measurements were observed, showing insufficient accuracy of the technique[51]. Some authors argued that the viability is not measured directly by fluorometry; it can also be overpredictive of nonviability and lead to unnecessary resections, particularly in patients with the isolated venous occlusion[34,50].

Recently a more sophisticated method, LFA, has been validated for intestinal microcirculation assessment. The method is based on intravenous injection of fluorescent dye (indocyanine green) and illumination of the bowel at the site of interest with a laser light. Digital videos of the fluorescence are recorded as a function of tissue perfusion[55]. The system is reliable and the measurements can be repeated after the rapid clearance of the fluorescent dye by hepatic uptake and biliary excretion. This is an advantage compared to the conventional perfusion fluorometry. In a study by Kudzsus et al[2], the predictive value of LFA was analyzed in a retrospective matched-pairs analysis. Measurements by LFA led to extended resections of malperfused intestine in 14.2% of patients, avoiding leaving nonviable bowel in situ. In 2% of the cases, resection of the viable bowel could be avoided. The use of intraoperative LFA reduced the risk of revision due to anastomotic leakage by 60% in patients undergoing elective colorectal surgery irrespective of their age and by 64% in patients above the age of 70. Hand-sewn anastomosis controlled by LFA reduced the risk of revision by 84% compared to the same anastomotic technique without LFA. The authors recommend the technique for routine use in clinical practice. However, the technique has limitations too: the contents of the intestine can interfere with fluorescence and the limiting values of LFA representing irreversible necrosis are not yet defined. The low LFA values were not consistent with pathological grade of ischemia in an animal model [50]. Moreover, the measurements are position dependent and not applicable for the assessment of colorectal anastomosis situated in the minor pelvis[1,3,56].

Thermal or infrared imaging is a scanning technique for recording small temperature differences between adjacent structures on a photographic display. The current medical application of thermal imaging is mainly limited to the detection of peripheral circulatory disturbances or breast cancer and to assess graft patency in plastic or cardiac surgery. Low surface temperature can be interpreted to be the poor vascularization that may cause an anastomotic impairment[57].

Thermal imaging was applied for assessment of the intestinal blood supply in several studies. The first study investigating a reactive bowel hyperemia after an ischemia and reperfusion was published in 1978[58]. Brooks et al[59] compared thermal imaging to visual inspection, Doppler ultrasound and fluorescence with Wood’s lamp. All methods had a high positive predictive value in detecting bowel ischemia, except the visual inspection which was the only method unable to detect a difference between vascularized and devascularized bowel. Doppler ultrasound and thermal imaging were 100% sensitive for a necrotic bowel[59]. Roberts et al showed that infrared imaging is a potential tool for localizing anatomic structures and assessing tissue viability during laparoscopic surgical procedures in a porcine model[60]. Nishikawa et al examined the use of intraoperative thermal imaging to assess gastric vascularization and gastric tube viability in patients during esophagectomy. They found intraoperative thermal imaging to be a noninvasive and reliable technique[57]. However, this is an indirect indicator of perfusion and oxygenation and the measurements are dependent on ambient temperature.

The principle of Laser Doppler flowmetry (LDF) is to measure the Doppler shift — the frequency change that light undergoes when reflected by moving objects such as red blood cells. LDF works by illuminating the tissue under observation with a monochromatic laser from a probe. When the tissue is illuminated, only 3%–7% of the light is reflected. The remaining 93%–97% is either absorbed by various structures or undergoes scattering. Another optical fiber collects the backscattered light from the tissue and returns it to the monitor. As a result, LDF produces an output signal that is proportional to the number of blood cells moving in the measured volume X mean velocity of the cells[61]. The measurements are expressed as mL/min per l00 g of tissue.

Vignali et al tested the reliability of intraoperative LDF measurements in predicting the occurrence of anastomotic leak in patients with colorectal cancer[62]. In a prospective study of 55 patients, the transmural colonic blood flow was measured during different stages of large bowel resection. Mean rectal stump flow reduction of 6.2% and proximal stump flow reduction of 5.1% was detected. The cut-off values of flow reduction, indicative of anastomotic insufficiency, were 16% and 12.9% respectively.

LDF measurements were also applied by Seike et al[5]. Colonic blood flow at the proximal site of the anastomosis was measured by LDF in 96 patients with cancer of the rectum and sigmoid colon while clamping inferior mesenteric artery (IMA) or left colic artery (LCA). Blood flow measured by LDF was significantly decreased by clamping the arteries. The flow reduction rate by IMA clamping was completely dependent on the individual patient and varied from zero to more than 80%. The study suggests that patients demonstrating significant decrease in blood flow by IMA clamping (more than 50%) may be at risk for the anastomotic ischemia. In such cases, the authors recommend preservation of LCA.

Nakatsuka measured gastric mucosal pH by a gastric tonometry and colonic tissue blood flow at the serosal layer by LDF during infrarenal abdominal aortic surgery in a prospective comparison study of eight patients[63]. In this study, the borderline value 41.7 ± 7.4 mL/min per l00 g of the sigmoid colonic tissue blood flow determined by LDF was detected which is sufficient to prevent postoperative ischemic colitis. Decrease of gastric mucosal pH measured by a gastric tonometry was not clinically significant and it did not reflect changes in intestinal microcirculation following aortic cross-clamping.

Singh and coworkers investigated VLS and LDF for assessing bowel serosal and mucosal oxygenation and blood flow in a pilot study of seven patients during different stages of colon surgery[64]. The authors found that the mean sigmoid mucosal StO₂ decreased after ligation of inferior mesenteric artery from the baseline of 73% to 55%. After complete devascularization, the mean sigmoid mucosal StO₂ was reduced to 39%. The mean sigmoid mucosal flux decreased after ligation of inferior mesenteric artery and it further decreased after complete devascularization. A similar trend was seen on the serosal flux measurements. In contrast, sigmoid serosal StO₂ measured by VLS did not decrease after ligation of inferior mesenteric artery and it decreased only slightly after complete devascularization from a baseline of 86% to 79%. The authors believe that mucosal StO₂ measurements by VLS can accurately diagnose bowel ischemia but serosal StO₂ measurements do not reflect a mucosal ischemia. Doppler flux measurements could be useful in detecting ischemia from the serosal surface of the bowel.

A major limitation of LDF is that it does not take into account the heterogeneity of blood flow as the velocity measurements represent the average of velocities in all vessels of the window studied.

LDF data are affected by motion artifacts in recording sites. In measurements of the oral mucosa, mucosa and serosa of the colon, large standard deviations of the flow parameters were observed[64,65]. Continuous measurement holding the probe securely may be necessary to overcome this limitation[5]. Pulsatile blood flow is required as the blood flow is measured in the deeper layers of the bowel containing arterioles. Changes in the patient’s cardiovascular status caused by intraoperative blood loss or drug administration through epidural anesthesia can also influence the measurements by LDF and might be time consuming since repeated measurements are necessary[5,21]. LDF requires tissue contact and may disturb local blood flow[1]. Scanning LDF was proposed for intraoperative use to overcome the limitations of single-point LDF[66,67]. In an animal study, this technique has been shown to have a significant relationship with histological grade of ischemia[68]. However, it is time consuming since the time needed to produce a perfusion image may be several minutes. The probability that movement artifacts will degrade the image quality increases with acquisition time[69]. The cut-off level of flow indicative of intestinal ischemia varies in different studies. In an animal study, a value of 30% of flow in the intact colon measured by LDF was considered to be safe for anastomotic healing[70]. In human studies, much lower decreases in LDF values showed risk of ischemia. The perfusion units are arbitrary and rather qualitative. Further studies are needed to create references in order to have quantitative values, the tools for distinction between capillary and global tissue perfusion and the tools providing measurements avoiding motion artifacts[69].

In 1986, the electronic contractility meter (ECM) was introduced as a quantitative method of measuring contractile activity of ischemic bowel. The ECM consists of two major components, a specially designed probe and an electronic control unit. The probe is clipped onto the serosal surface of the bowel and the electromyogram reflecting peristalsis of the small intestine is recorded at each 2 cm interval and quantified in millivolts by a computer algorithm[71]. Low electromyography (EMG) values referred to ischemic damage of the submucosal neuromuscular plexus[72].

Brolin et al have applied EMG, Doppler ultrasound and perfusion fluorometry. They showed that EMG might be a more sensitive indicator of ischemic damage than either ultrasonography or perfusion fluorometry in a model of acute intestinal ischemia[72,73]. However, after reperfusion, a slow increase of EMG values was observed and after 15 min normal values were reached. This shows the technique to be time consuming when no definitive necrosis of the intestine is present and the site of bowel resection must be determined. Also, the limit values of bowel wall contractility are not precise. Brolin et al[72] considered 50% of the contractility of a normal bowel to be sufficient for the anastomotic healing in a dog model. However, extrapolation to human conditions might be difficult considering that the bowels of different species are not equally resistant to ischemia. Electromyography has been used clinically in a study by Dutkiewicz et al[74]. Myoelectrical small bowel activity was measured intraoperatively in patients with and without small bowel ischemia. The diminished myoelectrical activity of the ischemic bowel wall was observed, which well correlated with histological changes[74]. It has not found a wider clinical acceptance, probably due to many unanswered questions and the complexity of the technique.

Tonometry is a noninvasive method to determine the intramucosal pH (pHi) in hollow organs which correlates with the oxygen supply to the mucosa. During mucosal hypoperfusion, acidosis develops, thereby resulting in a decrease of pHi. The tonometer is a silicone balloon catheter which is introduced into the lumen of the gut, then filled with isotonic sodium chloride solution and allows the free diffusion of carbon dioxide (CO₂). After an equilibration period, pCO₂ of the saline filling the balloon is proportionate to the intestinal mucosal pCO₂. This value and that of the arterial bicarbonate are then used in the Henderson-Hasselbach equation to calculate the intramucosal pH: pHi = 6.1 + log10 [(HCO₃)/(0.03 x pCO₂ X F)], in which F is a time-dependent correction factor and [HCO₃] is the actual bicarbonate concentration (mmol/L) of the arterial blood sample[75,76].

Kamiya and coworkers monitored pHi levels in the intestinal lumen using a tonometer intraoperatively and after the operation[75]. 35 patients who underwent free jejunal transfer following pharyngolaryngoesophagectomy for laryngeal, hypopharyngeal and cervical esophageal cancer were enrolled in the prospective observational study. The critical value of pHi < 7.10 was detected which leads to jejunal graft microcirculation disturbance and graft necrosis. It was shown that pHi measurement using a tonometer is useful for finding vascular problems in free jejunal grafts.

Millan, in a prospective study of 90 patients with colorectal resections, determined pHi levels at 24 and 48 h postoperatively and found that pHi<7.28 in the first 24 h postoperatively is associated with 22 times higher risk of anastomotic leak [76].

Despite being an accurate technique, tonometry is rather used for postoperative monitoring of bowel ischemia[76-78]. The necessity of leaving the catheters in situ may be considered a shortcoming of the technique.

Deeba et al[77] employed a momental intramural microdialysis to measure glucose and lactate levels in the bowel wall of specimens being resected intraoperatively and to monitor bowel ischemia in seven patients. After mobilization and before transecting any feeding vessel, a microdialysis probe was tunneled in the seromuscular layer of the left colon and fixed in place. Glucose and lactate concentrations were monitored until the bowel specimen was removed. After the transection of the feeding arteries, the glucose concentrations decreased whereas lactate concentrations increased leading to increase of lactate/glucose ratio. The study demonstrated the feasibility of using the microdialysis system in clinical practice. The authors suggest implanting the catheter near a bowel segment that is of questionable viability and to tunnel it outside the abdomen for postoperative bedside monitoring to alert the physician before ischemia manifests clinically.

Matsuo et al[79] evaluated the viability of a strangulated intestine by measuring its electrical properties in an experimental study. Correlation of changes in dielectric parameters with intestinal ischemia was found. Values suggesting the need for resection of the nonviable bowel were found; however, no intermediate value showing the possible recovery could be determined.

Horgan and Gorey defined the requirements of an ideal bowel viability test: 1) The technique must have ready availability, preferably in every operating theater dealing with abdominal emergencies; 2) The necessary equipment must not be cumbersome or require specialized personnel; 3) The method must be accurate with a minimum of false-negative results and, more importantly, few false positives; 4) The technique must be objective and reproducible; and 5) The method must be cost effective[80].

Despite numerous techniques are available as aforementioned, only several have been used in human studies over the last ten years (Table 1). There is no agreement as to which method is the most accurate and best applicable. Some of the techniques are applicable for postoperative follow-up of intestinal microcirculation (tonometry and rapid sampling microdialysis). Various parameters are measured by different methods and the most widely used are those reflecting oxygenation and perfusion, measured by VLS and LDF respectively. Considering that oxygenation and perfusion would probably have to be both evaluated intraoperatively, these two techniques seem to be the most promising. However, only one study has made comparison between the LDF and VLS and the oximetry has not been compared to other techniques of intestinal viability assessment[64]. No cut-off values of parameters indicative of intestinal ischemia have been determined yet. Several authors obtained different results applying the same technique. So far, there is no method that would fulfill all requirements stated by Horgan and Gorey.

It has been estimated that blood flow in the mucosal and submucosal layers accounts for 70% of total blood flow in intestinal tissues; therefore, microcirculation should be evaluated in mucosal and submucosal layer[81]. In other experimental studies using LDF and hydrogen gas clearance, a strong correlation between submucosal and subserosal blood flow was shown[38,82]. It was also confirmed in human studies[1,20,21,63,64,77]. Brolin et al[72] observed that in an ischemia model, extensive mucosal damage does not consistently result in bowel infarction and suggested that the measurements should be obtained from the serosal surface. This led to less invasive microcirculation assessment at the serosa of the intestine[39]. However, measurements performed by different authors and different systems yield unequal results. Karliczek et al have shown the reproducible measurements at the serosa of the resected bowel end whereas Singh et al have showed that the serosal StO₂ measurements are not sensitive to ischemia and are only possible at the mucosa. Moreover, different baseline values of serosal StO₂ were detected by the studies, 72.1 ± 9.0% and 86 ± 7.3% respectively[1,64]. The authors used different oximeter systems and the inter-device variability could possibly be the reason for the different results.

There are multiple risk factors predisposing stoma complications, including smoking, diabetes, grade of operating surgeon and emergency procedure. Necrosis is very important among other stoma complications. An insufficient perfusion and ischemia of the resected bowel end is one of the predisposing risk factors. In a prospective observational study of 97 patients with a newly created intestinal stoma, three patients developed necrosis and two experienced ischemia of their stoma[83]. In a nationwide prospective audit in the UK, the observed frequency of stoma necrosis was 8.7%[84]. The incidence of stoma ischemia and necrosis could be potentially diminished by objective intraoperative evaluation of stoma viability and timely surgical correction. However, it relies mainly on subjective intraoperative findings. Adequacy of blood supply at the enterostomy resection margin was assessed by Doppler ultrasound in a previous study[32]. In two studies, stoma microcirculation was assessed postoperatively. Boerma and coworkers compared sublingual and stoma microcirculation using orthogonal polarization spectral imaging technique in patients with abdominal sepsis[85]. The method has been validated previously for microcirculatory research in sepsis patients[86]. There was neither correlation between intestinal and sublingual blood flow, nor between intestinal and systemic hemodynamic parameters in abdominal sepsis patients on day one of sepsis and on day three this correlation appeared to be restored. These findings suggest that intestinal microcirculation cannot always be judged from the parameters of other microcirculatory beds or systemic hemodynamic parameters. Singh and Harrison investigated tissue oxygen saturation in end colostomy stomas with a minimum age of three months using a visible light spectroscopy. The normal oxygenation value of 77.6 ± 6.8 was detected. There were no significant diurnal variations in the stomal oxygenation values[87].

Düchs and Foitzik analyzed the influence of systemic hemodynamics on microcirculatory parameters. The authors used three techniques in parallel: an intravital bowel mucosal microscopy for the measurement of capillary blood flow, a spectroscopy for the measurement of Hb oxygen saturation in the tissue and a polarographic measurement of mucosal pO₂. Capillary blood flow of the bowel mucosa measured by intravital microscopy did not demonstrate insufficient oxygen supply of the tissue under systemic hypoxia. This was demonstrated by another two techniques. Comparing microcirculation measurements by the three methods, the authors observed that the extent to which hypovolemia and hypoxia influence these parameters is different. Moreover, a poor correlation between measurements by the different methods was found. In conclusion, the mean arterial pressure, blood gases and hematocrit must be analyzed to correctly interpret microcirculatory parameters [15].