Journal of Critical Care (2013) 28, 538.e9–538.e14

Severe abnormalities in microvascular perfused vesseldensity are associated to organ dysfunctions andmortality and can be predicted by hyperlactatemia andnorepinephrine requirements in septic shock patientsGlenn Hernandez MDa,b,⁎, E. Christiaan Boerma MD, PhD c, Arnaldo Dubin MD, PhDd,Alejandro Bruhn MD, PhDb, Matty Koopmans RNc, Vanina Kanoore Edul MDd,Carolina Ruiz MDb, Ricardo Castro MDb, Mario Omar Pozo MDd, Cesar Pedreros MDb,Enrique Veas MDb, Andrea Fuentealba MDb, Eduardo Kattan MDb,Maximiliano Rovegno MDb, Can Ince PhDb

aDepartment of Translational Physiology, Academic Medical Center, University of Amsterdam, Meibergdreef 9,1105 AZ Amsterdam, The NetherlandsbDepartamento de Medicina Intensiva, Pontificia Universidad Catolica de Chile, Marcoleta 367 Santiago Chile 8320000cDepartament of Intensive Care, Medical Center Leeuwarden, 8901 BR Leeuwarden, The NetherlandsdServicio de Terapia Intensiva, Sanatorio Otamendi y Miroli, Buenos Aires C1115AAB, Argentina

2

0h

Keywords:Septic shock;Lactate;Microcirculation;Norepinephrine;Mixed venous oxygensaturation;

Perfusion

AbstractPurpose: The aims of this study are to determine the general relationship of perfused vessel density (PVD) tomortality and organ dysfunctions and to explore if patients in the lowest quartile of distribution for thisparameter present a higher risk of bad outcome and to identify systemic hemodynamic and perfusion variablesthat enhances the probability of finding a severe underlying microvascular dysfunction.Materials and Methods: This is a retrospective multicenter study including 122 septic shock patientsparticipating in 7 prospective clinical trials on which at least 1 sublingual microcirculatory assessment wasperformed during early resuscitation.Results: Perfused vessel density was significantly related to organ dysfunctions and mortality, but this effectwas largely explained by patients in the lowest quartile of distribution for PVD (P = .037 [odds ratio {OR}, 8.7;95% confidence interval {CI}, 1.14-66.78] for mortality). Hyperlactatemia (P b .026 [OR, 1.23; 95%CI, 1.03-1.47]) and high norepinephrine requirements (P b .019 [OR, 7.04; 95% CI, 1.38-35.89]) increased the odds offinding a severe microvascular dysfunction.Conclusions: Perfused vessel density is significantly related to organ dysfunctions andmortality in septic shockpatients, particularly in patients exhibiting more severe abnormalities as represented by the lowest quartile ofdistribution for this parameter. The presence of hyperlactatemia and high norepinephrine requirements increases

⁎ Corresponding author. Departamento de Medicina Intensiva, Pontificia Universidad Catolica de Chile, Marcoleta 367 Santiago Chile 8320000. Tel.: +563543972.E-mail address: glennguru@gmail.com (G. Hernandez).

883-9441/$ – see front matter © 2013 Elsevier Inc. All rights reserved.ttp://dx.doi.org/10.1016/j.jcrc.2012.11.022

538.e10 G. Hernandez et al.

the odds of finding a severe underlying microvascular dysfunction during a sublingual microcirculatoryassessment.© 2013 Elsevier Inc. All rights reserved.

1. Introduction

Microcirculatory perfusion has been the subject ofextensive research over the last decade, especially in thecontext of sepsis [1–4]. Several sublingual microcirculatoryabnormalities have been described and linked to morbidityand mortality in septic shock patients [1,3,5]. More recentresearch has been focused on potential therapies to improvemicrocirculatory flow with some promising results [6–11].However, there are still fundamental aspects that need to beresolved before launching major controlled clinical trials.First, although several indices to evaluate sublingualmicrocirculation have been proposed [5], the clinicalrelevance of individual microcirculatory parameters is farfrom being well established [1–4]. This is particularly truefor perfused vessel density (PVD), a parameter representingfunctional capillary density. Second, no severity staging hasbeen proposed or validated for any of these microcirculatoryindices. Third, the relationship between systemic hemody-namics, global perfusion parameters, and microcirculatoryderangements is still controversial [6,8,12–14], and as aconsequence, it is not clear if macrohemodynamic orperfusion parameters can predict the status of microcircula-tory perfusion.

As a matter of fact, a significant correlation betweensystemic hemodynamic parameters and microcirculatoryindices was observed in septic shock patients early on afteremergency hospital admission [12]. Nevertheless, thisfinding has not been reproduced in intensive care unit(ICU)–based studies [1,6]. In the same line, the relationshipbetween microcirculatory derangements and metabolicperfusion-related parameters is controversial. Althoughsome studies report parallel changes in microcirculatoryflow and lactate during septic shock resuscitation [6,11],others have failed to find any correlation [13]. Therefore,from a bedside perspective, it is hard to identify which septicshock patients are more likely to present a microvasculardysfunction and thus constitute better candidates formicrocirculatory assessment and for potential inclusion inclinical trials.

To address these uncertainties, we conducted a clinicalstudy in a large multicenter cohort of patients with septicshock. Our aims were to determine the general relationshipof PVD to mortality and organ dysfunctions and to explore ifpatients in the lowest quartile of distribution for thisparameter presented a higher risk of bad outcome. Inaddition, we wanted to identify systemic hemodynamic andperfusion variables that enhance the probability of finding asevere underlying microvascular dysfunction.

2. Methods

2.1. Setting

We conducted a retrospective, cross-sectional study in 3academic centers (Chile, Argentina, and The Netherlands).Septic shock patients [15] enrolled in 7 prospective studiesevaluating microcirculation in different settings wereconsidered for inclusion [4,7,9,10,13,16,17]. Although themain objectives of these studies were markedly different,they shared the following aspects: (1) local institutionalreview boards approval with informed consent requirement(except in study no. 17 for which it was waived by theinstitutional review board); (2) one or more microcirculatoryassessments performed in parallel with macrohemodynamicand perfusion parameters within the first 24 hours of septicshock diagnosis; (3) microcirculatory image analysis per-formed by an experienced investigator; (4) image analysiscriteria based on a recent consensus [5]; (5) demographic,severity scores, and clinical data registered at baseline andfollow-up until hospital discharge or death; and (6) overallmanagement of septic shock based upon perfusion-drivenprotocols, as summarized below.

2.2. Background management

Management protocols shared the following principles:Resuscitation was aimed at restoring macrohemodynamicsand global perfusion. Normalization of lactate was the mainend point in addition to early aggressive source control.Initial fluid resuscitation was directed at correcting basichemodynamic parameters. Norepinephrine (NE) was usedas the sole vasopressor in patients with persistenthypotension after fluid loading. Additional invasive mon-itoring, vasoactive drugs, and mechanical ventilation weredecided on an individual basis and center preferences byattending physicians. Cardiac index and related parameterswere evaluated with a pulmonary artery catheter. Intravas-cular volume status was optimized following recentrecommendations if suitable [18]. Unresponsive patientsreceived different rescue therapies [10,19] observing center-specific variations.

2.3. Microcirculatory assessment

Sublingual microcirculation was evaluated as soon astechnically feasible during the first 24 hours of resuscitation.Whenever microcirculation was evaluated, mean arterial

538.e11Severe abnormalities in microvascular PVD

pressure (MAP), cardiac index, lactate, mixed venous O2

saturation (SvO2), and NE doses were registered in parallel.The earliest microcirculatory assessment per patient wasconsidered for the present study.

Sublingual microcirculation was assessed with orthogonalpolarization spectral imaging [4,13] or side dark stream fieldvideo microscopy [7,9,10,16,17] with a 5× lens (Microscan;Microvision Medical, Amsterdam, The Netherlands). Ateach time point, at least 5, 10 to 20 seconds, images wererecorded. After removing saliva and oral secretions, theprobe was applied over the mucosa at the base of the tongue.Special care was taken to avoid pressure artifacts, which wasverified by checking ongoing flow in larger microvessels(N50 mm).

2.4. Microcirculatory analysis

Each center followed the recommendations of a recentconsensus conference [5], which proposes that image analysisshould be at least depicted as measurements of perfusion(proportion of perfused vessels (PPV, %) and density (totalvessel density [TVD] [n/mm] and PVD [n/mm]).

A grid line consisting of 3 horizontal and 3 verticalequidistant lines was superimposed on the image. Totalvessel density was calculated as the number of vesselscrossing the lines divided by the total length of the lines.All vessels crossing the lines were counted and classifiedeither as perfused (continuous flow) and nonperfused (noflow or intermittent flow) vessels. Proportion of perfusedvessels was calculated as the number of vessels withcontinuous flow divided by the total number of vessels,multiplied by 100. Perfused vessel density, an estimate offunctional capillary density, can be calculated as TVD ×PPV. Microcirculatory indices were determined for largeand small vessels using a cutoff diameter for small vesselsof less than 20 μm. Only data regarding the small vesselsare reported.

Considering that centers estimated microcirculatory flowindex and heterogeneity with different methodologies, theseflow indices were not included in the present analysis. On theother hand, we decided to focus our study mainly in PVDbecause in our opinion, it is the single approach that morecomprehensively quantifies microvascular perfusion.

2.5. Statistical analysis

Because of nonnormal distribution of microcirculatoryvariables, nonparametric statistics were used. Results areexpressed as median and interquartile range or mean whenpertinent. Medians were compared by Mann-Whitney U test,and categorical data, with χ2 test. Association betweenmicrocirculatory and hemodynamic and perfusion parame-ters was assessed with Spearman ρ.

A logistic regression analysis was performed to determineif PVD was an independent predictor of hospital mortality,incorporating several variables such as center, Acute

Physiology and Chronic health Evaluation (APACHE) IIscore, NE dose, lactate, and PVD quartile. Proportion ofperfused vessels was excluded from the model to avoidoveradjustment because PVD is calculated as TVD × PPV.

A second logistic regression analysis was performed todetermine systemic and perfusion variables independentlyassociated to the probability of presenting severe underlyingmicrocirculatory abnormalities as represented by the lowestPVD quartile. Several variables such as age, sex, APACHEscore, basal and 24-hour Sequential Organ Failure Assess-ment (SOFA) scores, NE dose, cardiac index, lactate, andSvO2 values were included in the model.

SPSS software version 17.0 (Chicago, Ill) was used forstatistical calculations. P b .05 was considered as statisticallysignificant All reported P values are 2 sided. For logisticregression analyses, P values of .05 and .1 were used as entryand retention criteria, respectively.

3. Results

One hundred twenty-two patients were included (age, 65years [18-84]; APACHE II score, 21 [18-25]; basal SOFAscore 10 [7-12]; 24-hour SOFA score 9 [7-11]; hospitalmortality, 33%). Main septic sources were abdominal 44%,respiratory 29%, urinary tract 10%, and catheter related 5%.Microcirculatory assessment was performed in all casesduring the first 12 hours after shock onset. Hemodynamicand perfusion parameters during microcirculatory assess-ment were the following: MAP, 67 (61-72) mm Hg; NEdose, 0.15 (0.01-0.43) μg/kg per minute; lactate, 2.3 (1.3-4.5) mmol/L; cardiac index, 3.8 (3.1-4.9) L/min per squaremeter; and SvO2, 71% (66%-76%).

Among microcirculatory indices, nonsurvivors exhibitedsignificantly lower PVD and PPV values as compared withsurvivors (PVD 11.3 [7.4-13.8] vs 13.8 [10.2-16.8] n/mm;P b .01, and PPV 79.8% [70.1%-93.5%] vs 91.1% [80%-100%]; P b .05). No difference was found for TVD (14.8[12.9-19.5] vs 14.8 [11.3-19.8] n/mm; P = .55). Whendividing PVD into quartile, we found that the lowest PVDquartile was independently associated to mortality (P =.037 [odds ratio [OR], 8.7; 95% confidence interval [CI],1.14-66.78) (Fig. 1), together with lactate (P = .02 [OR,1.5; 95% CI, 1.2-1.9]) and NE dose (P = .002 [OR, 55.3;95% CI, 4.6-666]). The other tested variables includingAPACHE II score (P = .66 [OR, 1.0; 95% CI, 0.9-1.-1])were not associated to hospital mortality in this study.

To address potential clinical relevance of these results, wecompared the lowest with the 3 upper PVD quartile in severalclinical, hemodynamic, or perfusion variables. Patients in thelowest PVD quartile (b9.4 n/mm) exhibited higher severityscores, NE requirements, and lactate levels (Table 1; Fig. 2).

When addressing the relationship between hemodynamicand perfusion parameters with this microcirculatory index,we found that PVD was significantly correlated with NE

Fig. 1 Relation between PVD quartile and mortality (estimateprobability).

Fig. 2 Arterial lactate distribution according to PVD quartile. Asignificant difference between the lowest and any other PVDquartile was observed (analysis of variance, P = .001, Bonferronipost hoc).

538.e12 G. Hernandez et al.

requirements and lactate levels (Spearman ρ, P b .0001 forboth), but not with MAP, cardiac index, or SvO2 values.

Only lactate (P b .026 [OR, 1.23; 95% CI, 1.03-1.47]) andNE requirements (P b .019 [OR, 7.04; 95% CI, 1.38-35.89])predicted the worst PVD quartile.

4. Discussion

Perfused vessel density was significantly related to organdysfunctions and mortality in our cohort of septic shockpatients, but this effect was largely explained by patientsexhibiting more severe microcirculatory abnormalities asrepresented by the lowest quartile of distribution for PVD.The presence of hyperlactatemia and high NE requirements

Table 1 Comparison of lowest vs upper PVD quartile

Quartile First (n = 30) Second tofourth (n = 92)

P

Age (y) 56 (48-68) 63 (55-76) .22APACHE II score 23 (18-27) 20 (16-23) .02 ⁎

SOFA score 12 (10-14) 9 (7-11) .001 ⁎

SOFA 24 h 12 (8-16) 9 (7-11) .002 ⁎

MAP (mm Hg) 67 (63-73) 70 (64-77) .21NE dose (μg/kgper minute)

0.37 (0.16-0.72) 0.05 (0-0.13) .0001 ⁎

Lactate (mmol/L) 5.8 (2.3-9.5) 2.1 (1.2-3.4) .002 ⁎

SVO2 (%) 75 (67-80) 71 (66-75) .064Cardiac index(L/min persquare meter)

4.6 (3.2-5.5) 4.0 (3.5-4.9) .68

Values are expressed as median (interquartile range).⁎ Mann-Whitney U test.

increased the odds of finding a severe underlyingmicrovascular dysfunction during a sublingual microcircu-latory assessment.

A recent consensus conference proposed a semiquantita-tive approach to evaluate microcirculatory abnormalities [5],but very few studies so far have been aimed at establishingtheir individual specific clinical relevance or hierarchy. Ourdata suggest that, from a clinical perspective, PVD is relevantdue to its association to major outcomes. This finding makesphysiological sense because this parameter takes intoaccount both the diffusional and convective determinantsof tissue oxygenation at the microvascular level [5].However, outcome relationships were mainly evident inpatients with more severe derangements, and thus, a stagingof severity for proposed microcirculatory indices appears asimperative. This is particularly important for future researchbecause microcirculatory-oriented trials should be focusedon patients with severe abnormalities, as determined by somestaging system, to effectively have the potential to impactmorbidity or mortality.

Perfused vessel density has been only recently highlight-ed in some small clinical or experimental microcirculatorystudies involving cardiogenic and hemorrhagic shock orseptic patients [20–23]. Den Uil et a [20] performed aprospective study of sublingual microcirculatory assessmentin a cohort of patients with cardiogenic shock and found thatorgan failures recovered more frequently in patients withPVD greater than median (10.3 n/mm) and that thisparameter was associated to outcome. Peruski and Cooper[21] demonstrated a good correlation between PVD andmacrohemodynamic variables in a canine model of septicshock. Yeh et al [22] demonstrated a significant correlationbetween lactate and PVD in a small cohort of patientsadmitted to the ICU after general or thoracic surgery. Finally,Top et al [23] found persistent low PVD values in

538.e13Severe abnormalities in microvascular PVD

nonsurvivors of sepsis in pediatric intensive care. Our studyinvolves a large cohort of septic shock patients and providesadditional robust data concerning the relevance of PVD as anexpression of microvascular dysfunction. Values reported forPVD across the literature are quite variable and in the rangeof 5.3 to 23 n/mm [10,16,20–25]. Thus, our finding of PVDvalues less than 9.5 n/mm in the lowest distribution quartileis consistent with current literature.

In relation to our second objective, we found a significantassociation between high NE requirements and hyperlacta-temia, with microcirculatory perfusion as depicted by thePVD parameter. Moreover, the probability of finding a PVDvalue in the lowest quartile (b9.4 n/mm) was particularlyhigher in patients with hyperlactatemia greater than 4 mmol/land NE requirements greater than 0.2 μg/kg per minute,representing a more severe septic shock state. Thus, severeseptic shock patients could represent a more precise targetfor interventions.

The association between NE requirements and microcir-culatory abnormalities does not necessarily imply a cause-effect relationship. In a recent study by Dubin et al [16], therewas considerable variation in the individual responses to NEthat were strongly dependent on the basal condition of themicrocirculation. In fact, PVD improved in patients with analtered sublingual perfusion at baseline and decreased inpatients with a preserved baseline microvascular perfusion.On the other hand, when MAP decreases below anautoregulatory threshold of 60 to 65 mm Hg, organ perfusionbecomes pressure dependent, and consequently, the use of NEto increase MAP could improve tissue perfusion as wasrecently confirmed by Georger et al [26] assessing near-infrared spectroscopy-related parameters. High NE require-ments could also simply reflect a more profound andwidespread circulatory dysfunction involving also the micro-vascular level. However, regardless of the presence of a cause/effect relationship, an important message for clinicians is thatpatients with high NE requirements are much more likely tohave severe microcirculatory derangements. Therefore, thesepatients are particularly suitable to microcirculatory assess-ment and potential inclusion in clinical trials.

Another interesting aspect of our study is the lack ofassociation between systemic parameters such as MAP,cardiac index, and SvO2 and microcirculatory abnormalitiesin a large series of septic shock patients. The case of SvO2 iscounter-intuitive because impaired oxygen extraction capac-ity has been attributed to potential microcirculatory abnor-malities during septic shock [3]. Even more, a normal SvO2

in the presence of hyperlactatemia has been advocated as anindication for the use of vasodilators to improve microcir-culatory flow [14]. However, SvO2 is a poor indicator ofmicrovascular dysfunction: SvO2 can be high or low for thesame degree of microvascular shunting according to arecently proposed theoretical model [26]. Several previousexperimental and clinical studies have shown conflictingdata about the relationship between SvO2 and microvascularalterations [1,27–30].

There are several possible explanations to this fact. Severemitochondrial dysfunction could theoretically impair tissueO2 extraction capabilities, and this form of cytophatichypoxia could be associated to high SvO2 values and arelatively preserved microcirculatory flow. On the otherhand, there is a good correlation between systemic, perfusion(including SvO2), and microcirculatory parameters in thepre-resuscitated septic patient early on after hospitaladmission [12], and all parameters tend to improve inparallel after a fluid challenge in volume-responsive patientsas has been suggested by several studies [31–33]. However,in patients who evolve into a persistent circulatorydysfunction (as is the case of our study population),individual perfusion markers may change in unpredictableor even opposite directions as we discuss in recent review[8]. As an example, sedation and mechanical ventilation maynormalize central venous O2 saturation in most ICU patients(and by the way influencing SvO2), despite persistentongoing tissue hypoxia and microcirculatory derangements[34,35]. Several recent articles support the heterogeneity ofhemodynamic and perfusion profiles in persistent sepsis-related circulatory dysfunction [13,8,36,37].

In addition, our study design does not allow us to establish acause-effect relationship between hyperlactatemia and micro-circulatory abnormalities, although microcirculatory dysfunc-tion can clearly induce hypoxic hyperlactatemia. However,concurrent nonhypoxic causes for hyperlactatemia mayjeopardize a clear-cut interpretation [8]. We also believe thatdiscrepancies between our study and previous reports involvingthis relationship [4,6,8,13] may be probably explained bydifferent study designs concerning timing and number ofmicrocirculatory assessments or therapeutic interventions.

Our study has several limitations. First, patients wereenrolled from clinical trials with different aims and settings.However, all fulfilled fundamental requirements in terms ofbackground clinical management and microcirculatoryassessment. Second, the retrospective design does notallow us to evaluate dynamic changes in microcirculatoryand systemic parameters. Third, we did not includemicrocirculatory flow index due to intercenter methodolog-ical differences in analysis. Thus, we do not know if theseresults are applicable to that parameter. Fourth, in thisanalysis, microcirculatory assessments were performed atdifferent time points within the first 24 hours of resuscitation.This fact does not invalidate our results because global andmicrocirculatory parameters were evaluated simultaneouslyin each case. In contrast, we acknowledge that therelationship between a single assessment of microcirculatorystatus and outcome is remote and thus can only be consideredas hypotheses generating.

In conclusion, we think that our study provides interestingdata for clinical research. Perfused vessel density issignificantly related to organ dysfunctions and mortality inseptic shock patients, but this effect is largely explained bypatients exhibiting more severe microcirculatory abnormal-ities, as represented by the lowest quartile of distribution for

538.e14 G. Hernandez et al.

PVD. The presence of hyperlactatemia and high NErequirements increases the odds of finding a severeunderlying microvascular dysfunction during a sublingualmicrocirculatory assessment.

References

[1] De Backer D, Creteur J, Preiser JC, et al. Microvascular blood flow isaltered in patients with sepsis. Am J Respir Crit Care Med 2002;166:98-104.

[2] Sakr Y, Dubois MJ, De Backer D, et al. Persistent microcirculatoryalterations are associated with organ failure and death in patients withseptic shock. Crit Care Med 2004;32:1825-31.

[3] Ince C. The microcirculation is the motor of sepsis. Crit Care 2005;9(Suppl 4):S13-9.

[4] Boerma EC, van der Voort PH, Spronk PE, et al. Relationship betweensublingual and intestinal microcirculatory perfusion in patients withabdominal sepsis. Crit Care Med 2007;35:1055-60.

[5] De Backer D, Hollenberg S, Boerma EC, et al. How to evaluate themicrocirculation: report of a round table conference. Crit Care 2007;11:R101.

[6] De Backer D, Creteur J, Preiser J, et al. The effects of dobutamine onmicrocirculatory alterations in patients with septic shock areindependent of its systemic effects. Crit Care Med 2006;34:403-8.

[7] Dubin A, Pozo MO, Casabella CA, et al. Comparison of 6%hydroxyethyl starch 130/0.4 and saline solution for resuscitation ofthe microcirculation during the early goal-directed therapy of septicpatients. J Crit Care 2010;25:659.e1-8.

[8] Ospina-Tascon G, Neves AP, Occhipinti G, et al. Effects of fluids onmicrovascular perfusion in patients with severe sepsis. Intensive CareMed 2010;36:949-55.

[9] Boerma EC, Koopmans M, Konijn A, et al. Effects of nitroglycerin onsublingual microcirculatory blood flow in patients with severesepsis/septic shock after a strict resuscitation protocol: a double-blind randomized placebo controlled trial. Crit Care Med 2008;38:93-100.

[10] Ruiz C, Hernandez G, Godoy C, et al. Sublingual microcirculatorychanges during high-volume hemofiltration in hyperdynamic septicshock patients. Crit Care 2010;14:R170.

[11] Sakr Y, Chierego M, Piagnerelli M, et al. Microvascular response tored blood cell transfusion in patients with severe sepsis. Crit Care Med2007;35:1639-44.

[12] Trzeciak S, Dellinger RP, Parrillo JE, et al. Early microcirculatoryperfusion derangements in patients with severe sepsis and septicshock: relationship to hemodynamics, oxygen transport, and survival.Ann Emerg Med 2007;49:88-98.

[13] Boerma EC, Kuiper MA, Kingma WP, et al. Disparity between skinperfusion and sublingual microcirculatory alterations in severe sepsisand septic shock: a prospective observational study. Intensive CareMed 2008;34:1294-8.

[14] Jansen TC, van Bommel J, Schoonderbeek FJ, et al. Early lactate-guided therapy in intensive care unit patients: a multicenter, open-label, randomized controlled trial. Am J Respir Crit Care Med 2010;182:752-61.

[15] Levy MM, Fink MP, Marshall JC, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS International sepsis definitions conference. Crit Care Med2003;31:1250-6.

[16] Dubin A, Pozo MO, Casabella CA, et al. Increasing arterial bloodpressure with norepinephrine does not improve microcirculatory bloodflow: a prospective study. Crit Care 2009;13:R92.

[17] Hernandez G, Bruhn A, Castro R, et al. Persistent sepsis-inducedhypotension without hyperlactatemia: a distinct clinical and physio-

logical profile within the spectrum of septic shock. Crit Care Res Pract2012;2012:536852.

[18] Lamia B, Chemla D, Richard C, et al. Clinical review: interpretation ofarterial pressure wave in shock states. Crit Care 2005;9:601-6.

[19] Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign:international guidelines for management of severe sepsis and septicshock: 2008. Intensive Care Med 2008;34:17-60s.

[20] Den Uil CA, Lagrand WK, Van der Ent M, et al. Impairedmicrocirculation predicts poor outcome of patients with acutemyocardial infarction complicated by cardiogenic shock. Eur Heart J2010;31:3032-9.

[21] Peruski AM, Cooper ES. Assessment of microcirculatory changes byuse of sidestream dark field microscopy during hemorrhagic shock indogs. Am J Vet Res 2011;72:438-45.

[22] Yeh YC, Wang MJ, Chao A, et al. Correlation between earlysublingual small vessel density and late blood lactate level in criticallyill surgical patients. J Surg Res 2012 [Epub ahead of print].

[23] Top AP, Ince C, de Meij N, et al. Persistent low microcirculatoryvessel density in nonsurvivors of sepsis in pediatric intensive care. CritCare Med 2011;39:8-13.

[24] Jhanji S, Stirling S, Patel N, et al. The effect of increasing doses ofnorepinephrine on tissue oxygenation and microvascular flow inpatients with septic shock. Crit Care Med 2009;37:1961-6.

[25] Morelli A, Donati A, Ertmer C, et al. Levosimendan for resuscitatingthe microcirculation in patients with septic shock: a randomizedcontrolled study. Crit Care 2010;14:R232.

[26] Georger JF, Hamzaoui O, Chaari A, et al. Restoring arterial pressurewith norepinephrine improves muscle tissue oxygenation assessed bynear-infrared spectroscopy in severely hypotensive septic patients.Intensive Care Med 2010;36:1882-9.

[27] De Backer D, Ospina-Tascon G, Salgado D. Monitoring themicrocirculation in the critically ill patient: current methods andfuture approaches. Intensive Care Med 2010;36:1813-25.

[28] Farquhar I, Martin CM, Lam C, et al. Decreased capillary density invivo in bowel mucosa of rats with normotensive sepsis. J Surg Res1996;61:190-6.

[29] Goldman D, Bateman RM, Ellis CG. Effect of decreased O2 supply onskeletal muscle oxygenation and O2 consumption during sepsis: roleof heterogeneous capillary spacing and blood flow. Am J Physiol HeartCirc Physiol 2006;290:H2277-85.

[30] Textoris J, Fouche L, Wiramus S, et al. High central venous oxygensaturation in the latter stages of septic shock is associated withincreased mortality. Crit Care 2011;26:R176.

[31] Podbregar M, Mozina H. Skeletal muscle oxygen saturation does notestimate mixed venous oxygen saturation in patients with severe leftheart failure and additional severe sepsis or septic shock. Crit Care2007;11:R6.

[32] Payen D, Luengo C, Heyer L, et al. Is thenar tissue hemoglobin oxygensaturation in septic shock related to macrohemodynamic variables andoutcome? Crit Care 2009;13(Suppl 5):S6.

[33] Pottecher J, Deruddre S, Teboul JL, et al. Both passive leg raising andintravascular volume expansion improve sublingual microcirculatoryperfusion in severe sepsis and septic shock patients. Intensive CareMed 2010;36:1867-74.

[34] Hernandez G, Bruhn A, Castro R, et al. The holistic view on perfusionmonitoring in septic shock. Curr Opin Crit Care 2012;18:280-6.

[35] Hernandez G, Peña H, Cornejo R, et al. Impact of emergencyintubation on central venous oxygen saturation in critically ill patients:a multicenter observational study. Crit Care 2009;13:R63.

[36] Vallée F, Vallet B, Mathe O, et al. Central venous-to-arterial carbondioxide difference: an additional target for goal-directed therapy inseptic shock? Intensive Care Med 2008;34:2218-25.

[37] Lima A, van Bommel J, Sikorska K, et al. The relation of near-infraredspectroscopy with changes in peripheral circulation in critically illpatients. Crit Care Med 2011;39:1649-54.

Reproduced with permission of the copyright owner. Further reproduction prohibited withoutpermission.