Near-infrared spectroscopy-guided closed-loop resuscitation of hemorrhage

Closed-loop automated critical care as proof-of-concept study for resuscitation in a swine model of ischemia-reperfusion injury

Background

Volume expansion and vasopressors for the treatment of shock is an intensive process that requires frequent assessments and adjustments. Strict blood pressure goals in multiple physiologic states of shock (traumatic brain injury, sepsis, and hemorrhagic) have been associated with improved outcomes. The availability of continuous physiologic data is amenable to closed-loop automated critical care to improve goal-directed resuscitation.

Methods

Five adult swine were anesthetized and subjected to a controlled 30% estimated total blood volume hemorrhage followed by 30 min of complete supra-celiac aortic occlusion and then autotransfusion back to euvolemia with removal of aortic balloon. The animals underwent closed-loop critical care for 255 min after removal of the endovascular aortic balloon. The closed-loop critical care algorithm used proximal aortic pressure and central venous pressure as physiologic input data. The algorithm had the option to provide programmatic control of pumps for titration of vasopressors and weight-based crystalloid boluses (5 ml/kg) to maintain a mean arterial pressure between 60 and 70 mmHg.

Results

During the 255 min of critical care the animals experienced hypotension (< 60 mmHg) 15.3% (interquartile range: 8.6–16.9%), hypertension (> 70 mmHg) 7.7% (interquartile range: 6.7–9.4%), and normotension (60–70 mmHg) 76.9% (interquartile range: 76.5–81.2%) of the time. Excluding the first 60 min of the critical care phase the animals experienced hypotension 1.0% (interquartile range: 0.5–6.7%) of the time. Median intervention rate was 8.47 interventions per hour (interquartile range: 7.8–9.2 interventions per hour). The proportion of interventions was 61.5% (interquartile range: 61.1–66.7%) weight-based crystalloid boluses and 38.5% (interquartile range: 33.3–38.9%) titration of vasopressors.

Conclusion

This autonomous critical care platform uses critical care adjuncts in an ischemia–reperfusion injury model, utilizing goal-directed closed-loop critical care algorithm and device actuation. This description highlights the potential for this approach to deliver nuanced critical care in the ICU environment, thereby optimizing resuscitative efforts and expanding capabilities through cognitive offloading. Future efforts will focus on optimizing this platform through comparative studies of inputs, therapies, and comparison to manual critical care.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40635-022-00459-2.

Can central-venous oxygen saturation be estimated from tissue oxygen saturation during a venous occlusion test?

OBJECTIVE

To test whether tissue oxygen saturation (StO2) after a venous occlusion test estimates central venous oxygen saturation (ScvO2).

METHODS

Observational study in intensive care unit patients. Tissue oxygen saturation was monitored (InSpectra Tissue Spectrometer Model 650, Hutchinson Technology Inc., MN, USA) with a multiprobe (15/25mm) in the thenar position. A venous occlusion test in volunteers was applied in the upper arm to test the tolerability and pattern of StO2 changes during the venous occlusion test. A sphygmomanometer cuff was inflated to a pressure 30mmHg above diastolic pressure until StO2 reached a plateau and deflated to 0mmHg. Tissue oxygen saturation parameters were divided into resting StO2 (r-StO2) and minimal StO2 (m-StO2) at the end of the venous occlusion test. In patients, the cuff was inflated to a pressure 30mmHg above diastolic pressure for 5 min (volunteers' time derived) or until a StO2 plateau was reached. Tissue oxygen saturation parameters were divided into r-StO2, m-StO2, and the mean time that StO2 reached ScvO2. The StO2 value at the mean time was compared to ScvO2.

RESULTS

All 9 volunteers tolerated the venous occlusion test. The time for tolerability or the StO2 plateau was 7 ± 1 minutes. We studied 22 patients. The mean time for StO2 equalized ScvO2 was 100 sec and 95 sec (15/25mm probes). The StO2 value at 100 sec ([100-StO2] 15mm: 74 ± 7%; 25mm: 74 ± 6%) was then compared with ScvO2 (75 ± 6%). The StO2 value at 100 sec correlated with ScvO2 (15 mm: R2 = 0.63, 25mm: R2 = 0.67, p < 0.01) without discrepancy (Bland Altman).

CONCLUSION

Central venous oxygen saturation can be estimated from StO2 during a venous occlusion test.

Performance of closed-loop resuscitation in a pig model of haemorrhagic shock with fluid alone or in combination with norepinephrine, a pilot study

We evaluated the performance of a new device to control the administration of fluid alone or co-administration of fluid and norepinephrine in a pig model of haemorrhagic shock in two sets of experiments. In the first one, resuscitation was guided using continuous arterial pressure measurements (three groups: resuscitation with fluid by a physician, CL resuscitation with fluid, and CL resuscitation with fluid and norepinephrine). In the second one, resuscitation was guided using discontinuous arterial pressure measurements (three groups: CL resuscitation with fluid alone, CL resuscitation with fluid and moderate dose norepinephrine, and CL resuscitation with fluid and a high dose of norepinephrine). Pigs were resuscitated for 1 h. In the first set of experiments, proportion of time spent in the target area of 78-88 mmHg of systolic arterial pressure was not statistically different between the three groups: manual, 71.2% (39.1-80.1); CL with fluid, 87.8% (68.3-97.4); and CL with fluid and norepinephrine, 78.1% (59.2-83.6), p = 0.151. In the second set of experiments, performance of CL resuscitation with fluid or with combination of fluid and high or moderate dose of norepinephrine was not significantly different (p = 0.543 for time in target). Pigs resuscitated with norepinephrine required less fluid and had less haemodilution than pigs resuscitated with fluid alone. Performance of CL resuscitation using continuous arterial pressure measurement was not significantly different than optimised manual treatment by a dedicated physician. Performance of CL resuscitation was reduced with discontinuous arterial pressure measurements in comparison with continuous arterial pressure measurements.

Performance of closed-loop resuscitation of haemorrhagic shock with fluid alone or in combination with norepinephrine: an experimental study

Background

Closed-loop resuscitation can improve personalization of care, decrease workload and bring expert knowledge in isolated areas. We have developed a new device to control the administration of fluid or simultaneous co-administration of fluid and norepinephrine using arterial pressure.

Method

We evaluated the performance of our prototype in a rodent model of haemorrhagic shock. After haemorrhagic shock, rats were randomized to five experimental groups: three were resuscitated with fluid and two with co-administration of fluid and norepinephrine. Among groups resuscitated with fluid, one was resuscitated by a physician and two were resuscitated according to two different closed-loop algorithms. Among groups resuscitated with fluid and norepinephrine, one was resuscitated by a physician and the other one by the closed-loop device. The precision of arterial pressure during the resuscitation period was assessed using rising time, time passed in the target area and performance error calculations.

Results

Groups resuscitated with fluid had similar performances and passed as much time in the target area of 80–90 mmHg as the manual group [manual: 76.8% (67.9–78.2), closed-loop: 64.6% (45.7–72.9) and 80.9% (59.1–85.3)]. Rats resuscitated with fluid and norepinephrine using closed-loop passed similar time in target area than manual group [closed-loop: 74.4% (58.4–84.5) vs. manual: 60.1% (46.1–72.4)] but had shorter rising time to reach target area [160 s (106–187) vs. 434 s (254–1081)] than those resuscitated by a physician. Rats resuscitated with co-administration of fluid and norepinephrine required less fluid and had less hemodilution than rats resuscitated with fluid alone. Lactate decrease was similar between groups resuscitated with fluid alone and fluid with norepinephrine.

Conclusions

This study assessed extensively the performances of several algorithms for closed-loop resuscitation of haemorrhagic shock with fluid alone and with co-administration of fluid and norepinephrine. The performance of the closed-loop algorithms tested was similar to physician-guided treatment with considerable saving of work for the caregiver. Arterial pressure closed-loop guided algorithms can be extended to combined administration of fluid and norepinephrine.

Electronic supplementary material

The online version of this article (10.1186/s13613-018-0436-0) contains supplementary material, which is available to authorized users.

Control-oriented physiological modeling of hemodynamic responses to blood volume perturbation

This paper presents a physiological model to reproduce hemodynamic responses to blood volume perturbation. The model consists of three sub-models: a control-theoretic model relating blood volume response to blood volume perturbation; a simple physics-based model relating blood volume to stroke volume and cardiac output; and a phenomenological model relating cardiac output to blood pressure. A unique characteristic of this model is its balance for simplicity and physiological transparency. Initial validity of the model was examined using experimental data collected from 11 animals. The model may serve as a viable basis for the design and evaluation of closed-loop fluid resuscitation controllers.

Closed-Loop- and Decision-Assist-Guided Fluid Therapy of Human Hemorrhage

OBJECTIVES

We sought to evaluate the efficacy, efficiency, and physiologic consequences of automated, endpoint-directed resuscitation systems and compare them to formula-based bolus resuscitation.

DESIGN

Experimental human hemorrhage and resuscitation.

SETTING

Clinical research laboratory.

SUBJECTS

Healthy volunteers.

INTERVENTIONS

Subjects (n = 7) were subjected to hemorrhage and underwent a randomized fluid resuscitation scheme on separate visits 1) formula-based bolus resuscitation; 2) semiautonomous (decision assist) fluid administration; and 3) fully autonomous (closed loop) resuscitation. Hemodynamic variables, volume shifts, fluid balance, and cardiac function were monitored during hemorrhage and resuscitation. Treatment modalities were compared based on resuscitation efficacy and efficiency.

MEASUREMENTS AND MAIN RESULTS

All approaches achieved target blood pressure by 60 minutes. Following hemorrhage, the total amount of infused fluid (bolus resuscitation: 30 mL/kg, decision assist: 5.6 ± 3 mL/kg, closed loop: 4.2 ± 2 mL/kg; p < 0.001), plasma volume, extravascular volume (bolus resuscitation: 17 ± 4 mL/kg, decision assist: 3 ± 1 mL/kg, closed loop: -0.3 ± 0.3 mL/kg; p < 0.001), body weight, and urinary output remained stable under decision assist and closed loop and were significantly increased under bolus resuscitation. Mean arterial pressure initially decreased further under bolus resuscitation (-10 mm Hg; p < 0.001) and was lower under bolus resuscitation than closed loop at 20 minutes (bolus resuscitation: 57 ± 2 mm Hg, closed loop: 69 ± 4 mm Hg; p = 0.036). Colloid osmotic pressure (bolus resuscitation: 19.3 ± 2 mm Hg, decision assist, closed loop: 24 ± 0.4 mm Hg; p < 0.05) and hemoglobin concentration were significantly decreased after bolus fluid administration.

CONCLUSIONS

We define efficacy of decision-assist and closed-loop resuscitation in human hemorrhage. In comparison with formula-based bolus resuscitation, both semiautonomous and autonomous approaches were more efficient in goal-directed resuscitation of hemorrhage. They provide favorable conditions for the avoidance of over-resuscitation and its adverse clinical sequelae. Decision-assist and closed-loop resuscitation algorithms are promising technological solutions for constrained environments and areas of limited resources.

Low near infrared spectroscopic somatic oxygen saturation at admission is associated with need for lifesaving interventions among unplanned admissions to the pediatric intensive care unit

To investigate the association between low near infrared spectroscopy (NIRS) somatic oxygen saturation (<70%) at admission and the need for lifesaving interventions (LSI) in the initial 24 h of a PICU admission. Retrospective chart review of all unplanned admissions to the pediatric intensive care unit (PICU) with NIRS somatic oxygen saturation data available within 4 h of admission, excluding admissions with a cardiac diagnosis. LSI data were collected for the first 24 h after admission. Hemodynamic parameters, laboratory values, illness severity scores and diagnoses were collected. Included PICU admissions were stratified by lowest NIRS value in the first 4 h after admission: low NIRS (<70%) and normal NIRS (≥70%) groups. Rate of LSI from 4 h to 24 h was compared between the two groups. Association of LSI with NIRS saturation and other clinical and laboratory parameters was measured by univariate and multivariate methods. We reviewed 411 consecutive unplanned admissions to the PICU of which 184 (44%) patients underwent NIRS monitoring. A higher proportion of patients who underwent somatic NIRS monitoring required LSIs compared to those without NIRS monitoring (36.4 vs 5.7% respectively, p < 0.0001). The proportion of patients who required LSI was higher in the group with low NIRS (<70%) within the first 4 h compared to those with normal NIRS (≥70%) (77.1 vs 22.1%, p < 0.0001). Fluid resuscitation, blood products and vasoactive medications were the most common LSIs. Multivariable modeling showed NIRS < 70% and heart rate > 2SD for age to be associated with LSIs. ROC curve analysis of the combination of NIRS < 70% and HR >2SD for age had an area under the curve of 0.79 with 78% sensitivity and 76% specificity for association with LSI. Compared to the normal NIRS group, the low NIRS group had higher mortality (10.4 vs 0.7%, p = 0.005) and longer median hospital length of stay (2.9 vs 1.6 days, p < 0.0001). Low somatic NIRS oxygen saturation (<70%) in the first 4 h of an unplanned PICU admission is associated with need for higher number of subsequent lifesaving interventions up to 24 h after admission. Noninvasive, continuous, somatic NIRS monitoring may identify children at high risk of medical instability.

Microvascular response to transfusion in elective spine surgery

AIM

To investigate the microvascular (skeletal muscle tissue oxygenation; SmO₂) response to transfusion in patients undergoing elective complex spine surgery.

METHODS

After IRB approval and written informed consent, 20 patients aged 18 to 85 years of age undergoing > 3 level anterior and posterior spine fusion surgery were enrolled in the study. Patients were followed throughout the operative procedure, and for 12 h postoperatively. In addition to standard American Society of Anesthesiologists monitors, invasive measurements including central venous pressure, continual analysis of stroke volume (SV), cardiac output (CO), cardiac index (CI), and stroke volume variability (SVV) was performed. To measure skeletal muscle oxygen saturation (SmO₂) during the study period, a non-invasive adhesive skin sensor based on Near Infrared Spectroscopy was placed over the deltoid muscle for continuous recording of optical spectra. All administration of fluids and blood products followed standard procedures at the Hospital for Special Surgery, without deviation from usual standards of care at the discretion of the Attending Anesthesiologist based on individual patient comorbidities, hemodynamic status, and laboratory data. Time stamps were collected for administration of colloids and blood products, to allow for analysis of SmO₂ immediately before, during, and after administration of these fluids, and to allow for analysis of hemodynamic data around the same time points. Hemodynamic and oxygenation variables were collected continuously throughout the surgery, including heart rate, blood pressure, mean arterial pressure, SV, CO, CI, SVV, and SmO₂. Bivariate analyses were conducted to examine the potential associations between the outcome of interest, SmO₂, and each hemodynamic parameter measured using Pearson’s correlation coefficient, both for the overall cohort and within-patients individually. The association between receipt of packed red blood cells and SmO₂ was performed by running an interrupted time series model, with SmO₂ as our outcome, controlling for the amount of time spent in surgery before and after receipt of PRBC and for the inherent correlation between observations. Our model was fit using PROC AUTOREG in SAS version 9.2. All other analyses were also conducted in SAS version 9.2 (SAS Institute Inc., Cary, NC, United States).

RESULTS

Pearson correlation coefficients varied widely between SmO₂ and each hemodynamic parameter examined. The strongest positive correlations existed between ScvO₂ (P = 0.41) and SV (P = 0.31) and SmO₂; the strongest negative correlations were seen between albumin (P = -0.43) and cell saver (P = -0.37) and SmO₂. Correlations for other laboratory parameters studied were weak and only based on a few observations. In the final model we found a small, but significant increase in SmO₂ at the time of PRBC administration by 1.29 units (P = 0.0002). SmO₂ values did not change over time prior to PRBC administration (P = 0.6658) but following PRBC administration, SmO₂ values declined significantly by 0.015 units (P < 0.0001).

CONCLUSION

Intra-operative measurement of SmO₂ during large volume, yet controlled hemorrhage, does not show a statistically significant correlation with either invasive hemodynamic, or laboratory parameters in patients undergoing elective complex spine surgery.

Near-Infrared Spectroscopy: The New Must Have Tool in the Intensive Care Unit?

Standard hemodynamic monitoring such as blood pressure and pulse oximetry may only provide a crude estimation of organ perfusion in the critical care setting. Near-infrared spectroscopy (NIRS) is based on the same principle as a pulse oximeter and allows continuous noninvasive monitoring of hemoglobin oxygenation and deoxygenation and thus tissue saturation "StO2" This review aims to provide an overview of NIRS technology principles and discuss its current clinical use in the critical care setting. The study selection was performed using the PubMed database to find studies that investigated the use of NIRS in both the critical care setting and in the intensive care unit. Currently, NIRS in the critical care setting is predominantly being used for infants and neonates. A number of studies in the past decade have shown promising results for the use of NIRS in surgical/trauma intensive care units during shock management as a prognostic tool and in guiding resuscitation. It is evident that over the past 2 decades, NIRS has gone from being a laboratory fascination to an actively employed clinical tool. Even though the benefit of routine use of this technology to achieve better outcomes is still questionable, the fact that NIRS is a low-cost, noninvasive monitoring modality improves the attractiveness of the technology. However, more research may be warranted before recommending its routine use in the critical care setting.

The ebb and flow of fluid (as in resuscitation)

Since the early 1960's "resuscitation" following major trauma involved use of replacement crystalloid fluid/estimated blood loss in volumes of 3/1, in the ambulance, emergency room, operating room and surgical intensive care unit. During the past 20 years, MAJOR paradigm shifts have occurred in this concept. As a result hypotensive resuscitation with a view towards restriction of crystalloid, and prevention of complications has occurred. Improved results in both civilian and military environments have been reported. As a result there is new focus on trauma surgical involvement in all aspects of trauma patient management, focus on early aggressive surgical approaches (which may or may not involve an operation), and movement from crystalloid to blood, plasma, and platelet replacement therapy.

500 ml of blood loss does not decrease non-invasive tissue oxygen saturation (StO2) as measured by near infrared spectroscopy - A hypothesis generating pilot study in healthy adult women

Background

The goal when resuscitating trauma patients is to achieve adequate tissue perfusion. One parameter of tissue perfusion is tissue oxygen saturation (StO₂), as measured by near infrared spectroscopy. Using a commercially available device, we investigated whether clinically relevant blood loss of 500 ml in healthy volunteers can be detected by changes in StO₂ after a standardized ischemic event.

Methods

We performed occlusion of the brachial artery for 3 minutes in 20 healthy female blood donors before and after blood donation. StO₂ and total oxygenated tissue hemoglobin (O₂Hb) were measured continuously at the thenar eminence. 10 healthy volunteers were assessed in the same way, to examine whether repeated vascular occlusion without blood donation exhibits time dependent effects.

Results

Blood donation caused a substantial decrease in systolic blood pressure, but did not affect resting StO₂ and O₂Hb values. No changes were measured in the blood donor group in the reaction to the vascular occlusion test, but in the control group there was an increase in the O₂Hb rate of recovery during the reperfusion phase.

Conclusion

StO₂ measured at the thenar eminence seems to be insensitive to blood loss of 500 ml in this setting. Probably blood loss greater than this might lead to detectable changes guiding the treating physician. The exact cut off for detectable changes and the time effect on repeated vascular occlusion tests should be explored further. Until now no such data exist.

Thenar oxygen saturation measured by near infrared spectroscopy as a noninvasive predictor of low central venous oxygen saturation in septic patients

OBJECTIVE

To validate thenar oxygen saturation (StO(2)) measured by near-infrared spectroscopy (NIRS) as a noninvasive estimation of central venous saturation (ScvO(2)) in septic patients.

DESIGN

Prospective observational study.

SETTING

A 26-bed medical-surgical intensive care unit at a university-affiliated hospital.

PATIENTS

Patients consecutively admitted to the ICU in the early phase of severe sepsis and septic shock, after normalization of blood pressure with fluids and/or vasoactive drugs.

MEASUREMENTS

We recorded demographic data, severity score, hemodynamic data, and blood lactate, as well as ScvO(2), and StO(2) measured simultaneously on inclusion. Patients were divided into two groups according to ScvO(2) values: group A, with ScvO(2) < 70%, and group B, with ScvO(2) > or = 70%.

RESULTS

Forty patients were studied. StO(2) was significantly lower in group A than in group B (74.7 +/- 13.0 vs. 83.3 +/- 6.2, P 0.018). No differences in age, severity score, hemodynamics, vasoactive drugs, or lactate were found between groups. Simultaneously measured ScvO(2) and StO(2) showed a significant Pearson correlation (r = 0.39, P 0.017). For a StO(2) value of 75%, sensitivity was 0.44, specificity 0.93, positive predictive value 0.92, and negative predictive value 0.52 for detecting ScvO(2) values lower than 70%.

CONCLUSIONS

StO(2) correlates with ScvO(2) in normotensive patients with severe sepsis or septic shock. We propose a StO(2) cut-off value of 75% as a specific, rapid, noninvasive first step for detecting patients with low ScvO(2) values. Further studies are necessary to analyze the role of noninvasive StO(2) measurement in future resuscitation algorithms.