Secondary Logo

Journal Logo

Clinical Science Aspects

Increase of Perfusion Index During Vascular Occlusion Test is Paradoxically Associated With Higher Mortality in Septic Shock After Fluid Resuscitation: A Prospective Study

Menezes, Igor Alexandre Côrtes de; Cunha, Cláudio Leinig da; Junior, Hipolito Carraro; Luy, Alain Márcio

Author Information
doi: 10.1097/SHK.0000000000001217



Septic shock still causes high morbidity and mortality (1). Although the systemic hemodynamic management with antimicrobial therapy are still cornerstones of treatment, robust evidence has shown that monitoring of perfusion and microcirculation of non-vital peripheral tissues as sublingual mucosal, muscle, and skin can identify patients with poor prognosis even after normalization of systemic macrocirculation parameters (2–5). In fact, an “uncoupling” between macrocirculation and microcirculation commonly occurs in sepsis (6) and pathophysiology includes nitrosative/oxidative injuries, endothelial dysfunction, and vasomotor dysregulation (2, 6). The monitoring of peripheral perfusion, in addition to safety and noninvasiveness of methods (7), has also received special attention for survival prediction because the non-vital vascular beds are among the first to deteriorate and the last to be restored after resuscitation (8).

The endothelial dysfunction and vascular hyporesponsiveness are among the major microcirculatory disturbances in sepsis (9) and can be evaluated by a standardized vascular occlusion test (VOT). After brief arterial occlusion, an excess of blood flow occurs and leads to regional reactive hyperemia, which is accepted as an estimate of endothelial functionality and microvascular reserve of examined tissue (3, 4). The multiple mechanisms are organ-dependent and must be considered in the interpretation of the results of test (10). In septic patients during VOT, reactive hyperemia was clearly reduced when evaluated in skeletal muscle microcirculation (3) and nitric oxide (NO) dependent vessels (11) suggesting a link between severity of sepsis and degree of vascular damages assessed at bedside: the lower proportion of recruitable microvasculature and worse dysfunction of the microvascular endothelium (3–4, 11).

New oximeters calculate the Perfusion Index (PI) from pulsatile photoplethysmography signal and indirectly measure the perfusion variations (7), mostly of cutaneous microcirculation (12). As observed with other methods (2), hypoperfusion measured with PI was also associated with higher mortality in sepsis (5). Surprisingly, using a time-response analysis of PI during VOT, it was recently demonstrated the existence of a time-dependent hyperemic response in septic shock: a clear hyporesponsiveness only in early phase of test with a significant microvascular ischemic reserve in posterior phase, despite the peripheral hypoperfusion (13).

The aims of this study were to use the analysis of PI during VOT to verify the prognostic value of peripheral microvascular reserve in septic shock and investigate the role of adrenergic stimulation in this finding. It also aimed to verify if this analysis of PI could improve the predictive value of arterial lactate, a well-established prognostic factor in septic shock.


Patients and study design

All participants or their legal representatives provided written informed consent and the research was approved by the Research Ethics Committee of the Universidade Federal do Paraná (protocol: 685.344/2014). Therefore, the study protocol was in accordance with the national and international ethical norms on research with human beings (conformed to the Declaration of Helsinki). This prospective-observational study was conducted in two intensive care units (ICUs) within the Hospital de Clínicas/Federal University of Paraná, from January 2015 to October 2017. The study selected consecutive adult patients admitted with the diagnosis of septic shock or within 24 h after septic shock onset in patients previously admitted for other causes. According to internationally accepted consensus definitions at the time of beginning of study (14, 15), sepsis was defined based on clinical evidence of infection and two or more of the following: fever axial temperature greater than 38°C or hypothermia (axial temperature <36°C), tachycardia (heart rate > 90 beats per min), tachypnea (>20 breaths per minute) or need for mechanical ventilation, leukocytosis (>12 000 cells/mm3) or leukopenia (<4,000 cells/mm3), or a ratio of greater than 10% band cells to polymorphonuclear cells. Septic shock was defined as sepsis with hypotension (mean arterial pressure [MAP] < 65 mm Hg) and/or hypoperfusion represented by hyperlactatemia (>2.0 mmol/L, irrespective of blood pressure), even after initial volume expansion and requiring vasopressors.

Exclusion criteria

Other causes of shock; severe hepatopathy/coagulopathy (platelets ≤ 20,000/mm3, international normalizated ratio [RNI] > 2.0, or activated partial thromboplastin time- aPTT > 70 s), infective endocarditis, systemic sclerosis, and severe obstructive arterial disease. These criteria were chosen to reduce the risks of possible hemorrhagic and ischemic complications of procedures. Patients with cardiac arrhythmias or active/severe bleeding were also excluded from the protocol for assessing adrenergic stimulus (increase of vasopressors doses).

Study protocol

All patients received broad-spectrum antibiotic coverage. The local hemodynamic support was as follows: central venous pressure was 8 mm Hg to 12 mm Hg; MAP > 65 mm Hg, urine output > 0.5 mL/kg/h and the central venous oxygen saturation (ScvO2) > 70% (15). Initially, all patients received 30 mL/kg crystalloid fluid over 01 h. Fluid administration was continued until lack of response to passive-leg raising or no respiratory variations of inferior vena cava diameter (Samsung Medison Ultrasound; Seoul, Korea). If MAP remained< 65 mm Hg after fluid administration, a diagnosis of septic shock was made and norepinephrine was titrated to maintain MAP> 65 mm Hg. Intensivists were blinded for peripheral perfusion variables. Capillary refill time (CRF) was also evaluated and was considered prolonged if above 4.5 s (16). The patients were followed until 28 days of diagnosis of shock or discharge from hospital.


Information collected included demographic characteristics, diagnosis for admission and comorbidities, Acute Physiology and Chronic Health Evaluation II score (APACHE II), and Sequential Organ Failure Assessment (SOFA) score. Assessment of patients occurred within 24 h after admission in ICU with the diagnosis of shock or within 24 h after the onset of shock in patients previously admitted for other causes. All hemodynamic, metabolic, and PI variables were measured after fluid resuscitation.

Simultaneous blood gases were obtained from arterial and central venous catheters inserted in the radial artery and superior vena cava (via the jugular of the subclavian vein), respectively. The catheters were inserted at admission in ICU. Samples were taken in 3-mL heparinized syringe. Blood gas analysis and lactate concentration were determined using the GEM premier 3,000 Gasometer (Barcelona, Spain). ScvO2 was calculated from a sample from central venous catheter. The central venous-arterial blood carbon dioxide partial pressure difference (Pv-aCO2) was calculated as the difference between the partial pressures of central venous carbon dioxide and arterial carbon dioxide.

Reactive hyperemia measured with PI during VOT

Intensivists were blinded for PI variables. PI were measured by attaching a pulse oximeter probe (Masimo Radical, Masimo-Corp, Irvine, Calif). The same researcher performed all tests. All studies were performed after resuscitation and at least 1 h of hemodynamic stability (no change in vasopressor dose or fluid boluses) in a controlled room temperature (25°C). The pulse oximeter was placed on the index finger. The PI was measured for a period of 05 min (basal value) after signal stabilization. Subsequently, a sphygmomanometer cuff was rapidly inflated (within 05 s) around the homolateral arm, 30 mm Hg to 50 mm Hg above the systolic pressure to occlude the arterial flow for a period of 03 min (3, 11, 13).

Reactive hyperemia occurred on deflation of the cuff. The PI was determined every 15 s for a period of 05 min to create a curve of PI variation/delta (Δ PI) as a function of time. The Δ PI was calculated at each assessed time point using the following formula: 

Next, the delta of peak of PI (ΔPI Peak) and time to reach the peak (time to peak) were measured. Also, the mean of the delta of PI was determined between 0 and 60 s after cuff deflation (ΔPI0–60) and 60 to 120 s after cuff deflation (ΔPI60–120). The ΔPI0–60 and ΔPI60–120 were determined using the mean of 05 PI values, recorded every 15 s at each interval. These time intervals were specially chosen with the purpose of evaluating the phases of reactive hyperemia mainly generated by mechanosensitive mechanisms and metabolic factors, respectively (13, 17). The ΔPI peak, time to peak, ΔPI0–60, and ΔPI60–120 were compared between groups.

In predictive analysis, patients were divided into two groups using median value of ΔPI peak (50th percentile) as cut-off point: the group that had values of ΔPI peak below the median and the group which had values above the median. The survival was compared between groups.

Hyperlactatemia after resuscitation and reactive hyperemia measured with PI

Persistent hyperlactatemia after fluid resuscitation was diagnosed if arterial lactate levels were above 2.0 mmol/L (16). The patients were divided into four groups based on presence or absence of hyperlactatemia and based on ΔPI peak values (the highest or lowest values groups). The mortality rate was compared between groups.

Adrenergic stimulus and reactive hyperemia measured with PI

In order to evaluate the role of adrenergic stimulus on reactive hyperemia measured with PI, a correlation test was performed between ΔPI peak values and noradrenaline doses of each patient.

In addition, in a convenience sample (consecutively selected patients) and 30 min after VOT, the doses of noradrenaline were increased during 15 min to obtain an increase from baseline MAP around 10 mm Hg (maximum MAP = 100 mm Hg and maximum noradrenaline dose = 2 mcg/kg/min). Patients were allowed to stabilize for 30 min. After this period a new VOT was performed. Values of basal PI and ΔPI peak obtained before and after increase of noradrenaline doses were then compared for each patient.



  • 1. Survival curves in groups with the highest and lowest ΔPI peak values using the median of the sample as cut-off point (main outcome).
  • 2. Analysis of PI during VOT (ΔPI peak, time to peak, ΔPI0–60 and ΔPI60–120) in survivors and nonsurvivors.


  • 1. Mortality in groups with the highest and lowest ΔPI peak values and its relationship with the presence of hyperlactatemia after fluid resuscitation.
  • Correlation test between noradrenaline doses and ΔPI peak values.
  • 3. Effects of increasing doses of noradrenaline on basal PI values and ΔPI peak values in a subgroup of patients with septic shock.

Statistical analysis

Shapiro–Wilk test was used to test normalcy of the sample. Nonparametric values are expressed as medians/interquartile ranges and categorical variables are expressed as percentages. Mann–Whitney test and Fisher exact test (with Holm–Bonferroni correction if multiple comparisons) were used to determine the significance of differences of nonparametric and categorical variables, respectively. Survival curves were computed by the Kaplan–Meier method and compared by the log-rank test. Spearman correlation analysis was conducted to determine the relationship between Reactive Hyperemia values and Noradrenaline doses. Wilcoxon signed rank test was used to compare the values of Perfusion index and Reactive Hyperemia before and after increase of noradrenaline doses.

We used data from our preliminary study to estimate sample size (13) and the sample median as the cut-off point. In that study, we had observed a hazard ratio of 2.92 for mortality in patients with the highest ΔPI peak values compared with group with the lowest ΔPI peak values. Assuming a 28-day mortality around 30% in patients with lowest ΔPI peak values and estimating a hazard ratio of 2.15 in the predictive study, we determined that the enrollment of 106 patients would provide a power of 80% to show a mortality difference at an alpha level of 0.05. Additionally, a posteriori receiver operator (ROC) curve was generated. The survival curves were also plotted based in cut-off point based on ROC curve. Concerning convenience sample size of “adrenergic stimulus protocol,” it was chosen based on a previous study (18). The statistical program GraphPad Prism 3.02 was used for all analyses.

This study followed STROBE guidelines for reporting results.


There were included 106 septic shock patients after fluid resuscitation. The in-hospital mortality of septic shock was 47.1% (50/106) (Supplemental Digital Content 1, The clinical-demographic and hemodynamic data of the patients are listed in Table 1. Taken as a whole, these data describe a heterogeneous critically ill population, typical of sepsis. The nonsurvivor group had higher APACHE and SOFA scores, needed higher noradrenaline doses, and more often used vasopressin than survivors. The nonsurvivor group also had lower MAP, lower urine output, higher arterial lactate levels, and more often had prolonged CRF than survivors. There were no differences in the clinical data, infection source, C-reactive protein, and other hemodynamic parameters between the survivors and the nonsurvivors.

Table 1:
The demographic, clinical, and hemodynamic of the septic shock patients after fluid resuscitation

Figure 1 shows the representation of Δ PI after deflation of sphygmomanometer cuff and Table 2 shows the basal PI, the peak of PI (Δ PI peak), the time to peak, the mean of Δ PI measured at the first minute interval (ΔPI 0–60) and at second minute interval (ΔPI 60–120) of groups. The basal PI values were lower and Δ PI peak values were higher in nonsurvivor group although the peaks were reached slower in this group. In addition, the reactive hyperemia was statistically similar in both groups at the early/mechanosensitive phase represented by ΔPI 0–60. However, the reactive hyperemia were higher in nonsurvivor group at the latter/metabolic phase represented by ΔPI 60-120.

Fig. 1:
Representation of delta of perfusion index (ΔPI) during vascular occlusion test in patients with septic shock.
Table 2:
Basal values of perfusion and parameters of reactive hyperemia using PI in septic shock: time to peak PI, Delta Peak of PI (ΔPI Peak), the early/mechanosensitive phase of reactive hyperemia at first minute after cuff deflation (ΔPI 0–60), and the latter/metabolic phase of reactive hyperemia at second minute after cuff deflation (ΔPI 60–120)

On the analysis of outcomes (Fig. 2), the predetermined cut-off value of ΔPI peak that separated the groups (50th percentile) was 66%. The clinical-demographic and hemodynamic data of these two groups are listed in Supplemental Digital Content 2 ( The group with ΔPI peak values above the cutoff value (higher microvascular reserve) had a mortality hazard more than two times larger than that group with ΔPI peak values bellow the cutoff value. In addition, the mortality hazard was even higher when the cut-off value was determined by the posteriori ROC curve (Supplemental Digital Content 3,

Fig. 2:
Kaplan–Meier survival curves of patients with septic shock with the lowest and the highest ΔPI peaks during vascular occlusion test.

As shown in Figure 3, in patients without persistent hyperlactatemia after fluid resuscitation, there were not statistically differences between the groups with higher and lower microvascular reserve estimated with ΔPI peak. In addition, the mortality rate of patients with persistent hyperlactatemia and lower ΔPI peaks was not statistically different from patients without hyperlactatemia. On the contrary, the mortality rate of patients with persistent hyperlactatemia and higher ΔPI peaks was extremely high and statistically different from all the other groups.

Fig. 3:
Association between peaks of PI during vascular occlusion test (ΔPI peaks), persistent hyperlactatemia after fluid resuscitation and mortality in septic shock.

As shown in Figure 4A, there was positive correlation between ΔPI peaks and vasopressor doses. Additionally, in a subgroup of 10 patients, the doses of noradrenaline were increased and MAP increased until reaching the predetermined target in all patients [from 80 ± 7 mm Hg to 91 ± 6 mm Hg, P < 0.001; median of dose increase: 32% (23%–40%)]. In these patients, there were evident and statistically significant reduction in PI (Fig. 4B) and concomitant increase in ΔPI peak values (Fig. 4C).

Fig. 4:
The adrenergic stimulus and peripheral microvascular reserve measured with PI.


Robust evidence has shown the direct association between microcirculatory disturbances and prognosis in sepsis (2–6). When evaluated using VOT, the degree of microvascular hyporesponsiveness seems to contribute directly to hypoperfusion and disease severity (3, 4, 11). Thus, the finding of higher microvascular reserve could indicate lower microvascular damage and better prognosis in septic shock. Using an analysis of PI during VOT, our main findings surprisingly point out that, when assessed at fingertip, the microvascular reserve should have another bedside interpretation because: its nonutilization occurs even in the presence of hyperlactatemia; it reflects a higher adrenergic stimulus; and it is related to higher mortality. In addition, this new interpretation of peripheral microvascular reserve may contribute to development of future guided therapy.

Several studies in diverse non-vital organs have shown the persistent hypoperfusion after correction of systemic hemodynamics as a strong predictor of mortality in septic shock (2, 4, 5). Our results are in line with those findings demonstrating worse perfusion parameters (CRF and PI) in nonsurvivors. However, the role of vascular hyporesponsiveness in resulting peripheral hypoperfusion is much less understood. When evaluated in skeletal muscle microcirculation (3, 4) and NO-dependent vessels (11) a direct relationship between these parameters seemed apparent and intuitive.

Reactive hyperemia after VOT is commonly used to verify impaired vascular reactivity. The increase in blood flow following brief arterial occlusion, therefore, represents the proportion of recruitable capillaries, arterioles, and small arteries upon minimal flow delivery (3). Beyond this micro-hemodynamic aspect, other important issues should be taken into account in the assessment of VOT and can provide clues to understanding our apparently paradoxical results. First, the reactive hyperemia magnitude is a clear organ-dependent phenomenon (3, 10, 11) and depends on different metabolic pathways present in different tissues. While endothelial-derived NO strongly mediates conduit vessel response (19) and cyclooxygenase products influence the muscle vascular reactivity (20), the skin reactive hyperemia is mediated by other mechanisms including sensory nerves and hyperpolarizing factors (10). In fact, differential impact of human sepsis on skin microvascular reactivity was suggested by Engelberger et al. (21) who showed that endotoxemia selectively inhibits nitric-oxide-dependent reactivity while the postischemic hyperemia remains preserved.

In addition, some mediators act as hyperpolarizing factors in microcirculation and are considerably increased in sepsis including calcitonin gene-related peptide (22) and oxidative stress-derived hydrogen peroxide (23, 24). These hyperpolarizing factors are associated with poor prognosis (22, 23) and could explain the relationship between preserved microvascular reactivity and mortality in septic shock. An evaluation of these mediators and reactive hyperemia is needed to confirm this hypothesis.

Another important issue concerns to the phase of hyperemic response evaluated because early flow responses seem to be derived, mainly, by mechanosensitive mechanisms while shear-stress and metabolic factors affect late flow responses (17). These facts could explain the diverse previous results measured with near-infrared spectroscopy in sepsis (NIRS) that evaluates the beginning of the first hyperemic phase (3). We showed that there was no blood flow measured with PI in most patients up to 15 s after cuff release (13), period in which the NIRS analysis after VOT is usually done (3, 4). In addition, the PI seems not to show a statistically significant difference between survivors and nonsurvivors in the first 60 s after deflation (mechanosensitive phase of test) suggested by the similar values of ΔPI0–60. Nonetheless, higher values of ΔPI60–120 in non-survivors suggested a difference between groups in the “metabolic phase” of reactive hyperemia.

It is well known that an intense redistribution of blood flow from non-vital organs to vital organs characteristically occurs in shock states causing peripheral hypoperfusion (25) which is generated primarily by an increase in sympathetic-neurohumoral activity (25, 26). Since previous reports showed that PI is a very sensitive method for adrenergic responses (27, 28) and we show a positive correlation between noradrenaline doses and ΔPI peak values, the adrenergic stimulus on the microcirculation (secondary to sympathetic response and to vasopressors) becomes a direct hypothesis that could explain the association between microvascular reserve and high mortality.

Nevertheless, previous clinical reports consistently showed that reactive hyperemia after VOT tends to be reduced, and not increased, by direct α-adrenergic stimulus when measured in cutaneous circulation (29). Similar results occurred in conductance vessels (30) and muscle sympathetic nerves (31). Another interesting fact is that sepsis impairs the adrenergic signaling and generates receptor hyporesponsiveness (9). Thus, requirement of increasing doses of vasopressors could reflect adrenergic desensitization. Therefore, to better verify the role of direct adrenergic stimulus, we evaluated the effects of noradrenaline increase in PI and ΔPI peak values for each patient. To ensure an effective pharmacologic response to drug, a dose increase was chosen up to predetermined target in the MAP (10 mm Hg). Using this protocol, it was demonstrated a reduction in PI and concomitant increase in ΔPI peak values. Therefore, our results strongly suggest that, at least in part, the adrenergic stimulus is related to the preserved peripheral microvascular reserve assessed with PI in septic shock.

Arterial lactate is often measured during the management of septic shock because of its recognized value as a predictor of mortality and as a marker of response to treatment (32). To evaluate whether prediction value of microvascular reserve measured with PI was affected by that parameter, we divided the patients based on arterial lactate levels and ΔPI peak after VOT. Our results suggest that without evidence of hyperlactatemia, increased ΔPI peak appears not to predict higher mortality. However, when persistent hyperlactatemia was present, ΔPI peak had shown to improve the predictive value: patients with lower ΔPI peak values and hyperlactatemia had mortality rate very close to those with normolactatemia; conversely, those with highest ΔPI peak values and concomitant hyperlactatemia had extremely high mortality rate.

It is accepted that lactate mainly arises from tissue hypoxia, when oxygen delivery fails to meet oxygen requirements, thus causing anaerobic glycolysis (32, 33). However, it also arises from other pathways as impaired oxygen extraction, increased glycolytic flux, and diminished lactate clearance (32). Concerning tissue hypoxia, hypoperfusion has been traditionally considered its most common cause (32, 33). However, it only correlates with abnormal macrohemodynamic during the initial phase of shock (33). There is increasing evidence that non-hypoxic and non-flow-dependent mechanisms influence the time course of lactate recovery rate and the distinction between these two scenarios (flow-dependent vs. non-flow-dependent hyperlactatemia) could strongly impact the management (34). Considering our results and phase of shock adopted for this study (after fluid resuscitation), the microvascular reserve evaluated with PI seems a promising parameter to differentiate these different clinical scenarios.

Finally, the identification of potential nonsurvivors for more precise therapy persists as an attractive idea in sepsis. Our results have provided some new perspectives to a future guided management. PI-VOT is an easy-to-use bedside test and early identifies at least two groups with different mortality risk. In addition, despite the microvascular structural damage in septic shock, the “functionality” of peripheral hypoperfusion was evidenced by the ischemic reserve in nonsurvivors and reinforces the need to study the impact of microvascular recruitment in these patients. Also, the potential improvement in the prediction of arterial lactate could theoretically increase the precision: it could avoid fluid overloading and/or excessive vasoactive dosage in hyperlactatemic patients with low mortality risk, increasing therapeutic efforts only in those with high mortality risk. Lastly, the relationship between PI-VOT and the adrenergic stimulation could contribute to monitoring of the sympathetic modulation, a treatment of recent interest for improving prognosis in human sepsis (35). However, the data reported in this study only allow assumptions about all these perspectives and further research is needed to verify them.

This study had limitations. First, this was a monocentric study with high mortality rate (tertiary hospital). Thus, a bigger multicenter study is needed to confirm these findings, also using a multivariate analysis to verify the value of PI-VOT as an independent prognostic factor. Second, a single measure was obtained, limiting conclusions about intraindividual reproducibility. A temporal evolution study is needed to verify the potential of delta-PI as dynamic parameter. Third, although our results have demonstrated the relationship of adrenergic stimulus to preserved microvascular reserve assessed with PI, they are not adequate to differentiate sympathetic response to shock and direct adrenergic effects of vasopressors. Finally, this study used only ΔPI peak values to calculate the sample size. Further research is needed to confirm the results using ΔPI and arterial lactate, concomitantly, as primary outcomes.

In conclusion, oximetry is widely prevalent in medical care and currently calculates the PI from pulsatile signal. We found that the increase of PI during VOT in septic shock denotes higher mortality and is explained, at least in part, by adrenergic stimulus. In addition, it was shown that assessment of PI-VOT appears to improve the predictive value of arterial lactate. We hypothesize that other mechanisms could also be responsible for these findings as hyperpolarizing factors. However, further investigations are necessary to clarify these assumptions.


1. Hotchkiss RS, Moldawer LL, Opal SM, Reinhart K, Turnbull IR, Vincent JL. Sepsis and septic shock. Nat Rev Dis Primers 2:16045, 2016.
2. De Backer D, Cortes DO, Donadello K, Vincent JL. Pathophysiology of microcirculatory dysfunction and the pathogenesis of septic shock. Virulence 5:73–79, 2014.
3. Doerschug KC, Delsing AS, Schmidt GA, Haynes WG. Impairments in microvascular reactivity are related to organ failure in human sepsis. Am J Physiol Heart Circ Physiol 293:H1065–H1071, 2007.
4. Shapiro NI, Arnold R, Sherwin R, O’Connor J, Najarro G, Singh S, Lundy D, Nelson T, Trzeciak SW, Jones AE. Emergency Medicine Shock Research Network (EMShockNet). The association of near-infrared spectroscopy-derived tissue oxygenation measurements with sepsis syndromes, organ dysfunction and mortality in emergency department patients with sepsis. Crit Care 15:R223, 2011.
5. He H, Liu D, Long Y, Wang XT. The peripheral perfusion index and transcutaneous oxygen challenge test are predicitive of mortality in septic patients after resuscitation. Crit Care 17:R116, 2013.
6. Hernandez G, Teboul JL. Is the macrocirculation really dissociated from the microcirculation in septic shock? Intensive Care Med 42:1621–1624, 2016.
7. Van Genderen ME, Van Bommel J, Lima A. Monitoring peripheral perfusion in critically ill patients at the bedside. Curr Opin Crit Care 18:273–279, 2012.
8. Lima A, Bakker J. Clinical monitoring of peripheral perfusion: there is more to learn. Crit Care 18:113, 2014.
9. Levy B, Collin S, Sennoun N, Daucrocq N, Kimmoun A, Asfar P, Perez P, Meziani F. Vascular hyporesponsiveness to vasopressors in septic shock: from bench to bedside. Intensive Care Med 36:2019–2029, 2010.
10. Roustit M, Cracowski J. Assesment of endothelial and neurovascular function in human skin microcirculation. Trends Pharmacol Sci 34:373–384, 2013.
11. Davis JS, Yeo TW, Thomas JH, McMillan M, Darcy CJ, McNeil YR, Cheng AC, Celermajer DS, Stephens DP, Anstey NM. Sepsis-associated microvascular dysfunction measured by peripheral arterial tonometry: an observational study. Crit Care 13:R155, 2009.
12. Shelley KH. Photoplethysmography: beyond the calculation of arterial oxygen saturation and heart rate. Anesth Analg 105:S31–36, 2007.
13. Menezes IA, Cunha CL, Junior HC, Luy AM. Perfusion index for assessment of microvascular reactivity in septic shock after fluid resuscitation. Rev Bras Ter Intensiva 30:135–143, 2018.
14. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 31:1250–1256, 2003.
15. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock. Intensive Care Med 39:165–228, 2013.
16. Dünser MW, Takala J, Brunauer A, Bakker J. Re-thinking resuscitation: leaving blood pressure cosmetics behind and moving forward to permissive hypotension and a tissue perfusion-based approach. Crit Care 17:326, 2013.
17. Koller A, Bagh Z. On the role of mechanosensitive mechanisms eliciting reactive hyperemia. Am J Physiol Heart Circ Physiol 283:H2250–2259, 2002.
18. Kienbaum P, Prante C, Lehmann N, Sander A, Jalowy A, Peters J. Alterations in forearm vascular reactivity in patients with septic shock. Anaesthesia 63:121–128, 2008.
19. Barac A, Campia U, Panza JA. Methods for evaluating endothelial function in humans. Hypertension 49:748–760, 2007.
20. Ardor G, Delachaux A, Dischl B, Hayoz D, Liaudet L, Waeber B, Feihl F. A comparative study of reactive hyperemia in human forearm skin and muscle. Physiol Res 57:685–692, 2008.
21. Engelberger RP, Pittet YK, Henry H, Delodder F, Hayoz D, Chioléro RL, Waeber B, Liaudet L, Berger MM, Feihl F. Acute endotoxemia inhibits microvascular nitric oxide dependent vasodilation in humans. Shock 35:28–34, 2011.
22. Beer S, Weinghardt H, Emmanuilidis K, Harzenetter MD, Matevossian E, Heidecke CD, Bartels H, Siewert JR, Holzmann B. Systemic neuropeptide levels as predictive indicators for lethal outcome in patients with postoperative sepsis. Crit Care Med 30:1794–1798, 2002.
23. Andrades ME, Morina A, Spasic S, Spasojević I. Bench-to-bedside review: sepsis-from the redox point of view. Crit Care 15:230, 2011.
24. Shimokawa H, Morikawa K. Hydrogen peroxide is an endothelium-derived hyperpolarizing factor in animals and humans. J Mol Cell Cardiol 39:725–732, 2005.
25. Miyagatani Y, Yukioka T, Ohta S, Ohta S, Matsuda H, Shimazu H, Shimazaki S. Vascular tone in patients with hemorrhagic shock. J Trauma 47:282–287, 1999.
26. Lima A, Jansen TC, Van Bommel J, Ince C, Bakker J. The prognostic value of the subjective assessment of peripheral perfusion in critically ill patients. Crit Care Med 37:934–938, 2009.
27. Kus A, Gurkan Y, Gormus SK, Solak M, Toker K. Usefulness of perfusion index to detect the effect of brachial plexus block. J Clin Monit Comput 27:325–328, 2013.
28. Rasmy I, Mohamed H, Nabil N, Abdalah S, Hasanin A, Eladawy A, Ahmed M, Mukhtar A. Evaluation of perfusion index as a predictor of vasopressor requirement in patients with severe sepsis. Shock 44:554–559, 2015.
29. Cankar K, Finderle Z, Strucl M. The effect of α-adrenoceptor agonists and L-NMMA on cutaneous postocclusive reactive hyperemia. Microvasc Res 77:198–203, 2009.
30. Thijssen DH, Atkinson CL, Ono K, Sprung VS, Spence AL, Pugh CJ, Green DJ. Sympathetic nervous system activation, arterial shear rate, and flow-mediated dilation. J Appl Physiol 116:1300–1307, 2014.
31. Sverrisdóttir YB, Jansson LM, Hagg U, Gan LM. Muscle sympathetic nerve activity is related to a surrogate marker of endothelial function in healthy individuals. PLoS One 5:e9257, 2010.
32. Suetrong B, Walley KR. Lactic acidosis in sepsis: it's not all anaerobic: implications for diagnosis and management. Chest 149:252–261, 2016.
33. Kiyatkin ME, Bakker J. Lactate and microcirculation as suitable targets for hemodynamic optimization in resuscitation of circulatory shock. Curr Opin Crit Care 23:348–354, 2017.
34. Hernandez G, Luengo C, Bruhn A, Kattan E, Friedman G, Ospina-Tascon GA, Fuentealba A, Castro R, Regueira T, Romero C, et al. When to stop septic shock resuscitation: clues from a dynamic perfusion monitoring. Ann Intensive Care 4:30, 2014.
35. Ferreira JA, Bissel BD. Misdirected Sympathy: the role of sympatholysis in sepsis and septic shock. J Intensive Care Med 33:74–86, 2017.

Microcirculation; mortality; oximetry; perfusion index; shock

Supplemental Digital Content

Copyright © 2018 by the Shock Society