Among both military and civilian trauma patients, hemorrhage is a leading cause of mortality (1,2) (1,2). Both early detection and management of blood loss by emergency first responders could potentially save thousands of lives annually. However, compensatory physiologic mechanisms allow individuals to maintain traditional vital signs within relatively normal ranges until severe cardiovascular compromise has occurred (3–5) (3–5) (3–5); thus, timely detection of blood loss with these traditional physiologic measures remains inadequate. Often the first clinical sign of the onset of hemodynamic decompensation following life-threatening blood loss is an acute and dramatic decline in arterial blood pressure which occurs immediately before collapse (6). Therefore, a noninvasive measure indicating the degree of blood loss well in advance of this decline in blood pressure would greatly improve the ability of medical providers to identify patients approaching the early phase of decompensatory shock.
Several indicators of blood loss have been proposed such as pulse pressure variability (PPV) (7), peripheral perfusion index (PPI) (8), and the recently described compensatory reserve index (CRI) (9–11) (9–11) (9–11). Data from a recent publication reported that during a controlled voluntary hemorrhage of 20% to 25% in 20 patients, the CRI had superior discriminative ability to predict blood loss when compared with standard vital signs (e.g., blood pressures, heart rate, arterial oxygen saturation) and traditional hemodynamic responses (e.g., stroke volume, cardiac output, systemic peripheral resistance) (9). More importantly, the CRI has consistently proven to display high discriminative ability for early prediction of the onset of hemodynamic decompensation in humans exposed to progressive reductions in central circulating blood volume (10) similar to that described in hemorrhage (12,13) (12,13). However, the ability of PPV and PPI to predict the onset of hemodynamic decompensation during progressive reductions in central circulating blood volume has not been investigated. As such, the purpose of this study was to compare the sensitivity and specificity of PPV, PPI, and CRI in predicting progression to the onset of hemodynamic decompensation. Compared with both PPI and PPV values, we hypothesized that the CRI response would demonstrate superior discriminative ability for early prediction of the onset of hemodynamic decompensation.
MATERIALS AND METHODS
Study participants and ethical approval
There were 51 healthy, normotensive, nonsmoking adults who volunteered to participate in this study that was conducted in the Human Physiology Laboratory at the US Army Institute for Surgical Research, Fort Sam Houston, Texas. This sample included both men (n = 28) and women (n = 23) with the following demographic characteristics (mean ± SE): age (25 ± 1 years), height (170.9 ± 0.6 cm), weight (73 ± 2.5 kg). The study was approved by the Institutional Review Board of the Brooke Army Medical Center, Fort Sam Houston, Texas (BAMC#H-11-038). All participants were instructed to maintain a normal sleep pattern and refrain from the following for 24 h before testing: exercise, alcohol consumption, stimulants (e.g. caffeine), decongestants, and other nonprescription drugs. Participants received both verbal and written summaries of the study procedures, which included the potential risks associated with the study. In addition, participants were familiarized with the protocol and instrumentation before participation. On the day of the study, each participant completed both a medical history and physical screening by a physician to officially be cleared to begin the study. Female participants underwent a urine pregnancy test within an hour of the start of the experiment to rule out pregnancy. Each participant gave written informed consent to participate in the study.
Study design and lower body negative pressure (LBNP) protocol
We used a one-group, pre- to post-treatment study design to determine the discriminative ability of PPI, PPV, and CRI to predict the onset of hemodynamic decompensation. To progressively simulate hemorrhage in conscious humans, central hypovolemia was induced by application of LBNP. Participants were instrumented for standard lead II electrocardiogram to record heart rate (HR); a finger photoplethysmography cuff (ccNexfin, Edwards Lifesciences, Irvine, CA) placed on the middle finger of the left hand to record beat-by-beat finger systolic, mean, and diastolic arterial pressures (SAP, MAP, DAP); and a standard pulse oximeter (Rainbow SET, Masimo Corp., Irvine, CA) placed on the middle finger of the right hand. Participants were then positioned supine within an airtight chamber that was sealed at the level of the iliac crest by a neoprene skirt, and allowed to acclimate to this position for 10–20 min before initiating the experiment. The LBNP protocol consisted of a 5-min control period (baseline) followed by stepwise chamber decompression of −15, −30, −45, −60, −70, −80, −90, and −100 mmHg. LBNP was increased every 5 min until the onset of hemodynamic decompensation.
Induction of central hypovolemia in humans using LBNP is known to result in the activation of compensatory physiologic mechanisms that are comparable to those seen during actual hemorrhage (9). The LBNP procedure used by our laboratory has previously been described in detail (13), and is an effective way to assess compensatory responses to progressive central hypovolemia in real time.
High versus low tolerant participants
Consistent with previous publications (14–16) (14–16) (14–16), participants were categorized as either high tolerant (HT) or low tolerant (LT) based on the level of LBNP at which they experienced the onset of hemodynamic decompensation. A participant was categorized as LT if he or she experienced hemodynamic decompensation before or during LBNP of −60 mmHg, and as HT if he or she experienced hemodynamic decompensation after LBNP of −60 mmHg.
The onset of hemodynamic decompensation was defined by at least one of the following criteria: sudden bradycardia, a precipitous fall in systolic pressure greater than 15 mmHg, progressive fall of systolic pressure below 80 mmHg, and/or voluntary termination due to the onset of presyncopal symptoms. These presyncopal symptoms included sweating, nausea, dizziness, vision alterations, or general discomfort.
Compensatory reserve index (CRI)
The CRI has recently been introduced by our laboratory as a novel machine-learning model that integrates data from the arterial waveforms of pulse oximeters to estimate how close an individual is to hemodynamic decompensation during progressive central hypovolemia (10). Detailed methodology for determining CRI has been previously described by Convertino et al. (10). Briefly, feature extraction and machine-learning techniques were used to collectively process the same analog photoplethysmographic (PPG) arterial waveforms obtained from the Masimo pulse oximeter that were used to calculate PPI and PPV during LBNP protocols. The initial predictive CRI models were developed from processing PPG arterial waveforms obtained from an infrared finger blood pressure cuff (10,17) (10,17). This original algorithm was transformed to interpreting PPGs obtained from the Masimo oximeter by conducting the same progressive LBNP protocols on 10 subjects for comparison of PPG features obtained from blood pressure waveforms to those obtained from flow (oximeter) waveforms. In this way, data analysis was designed to compare CRI, PPI, and PPV values derived from the same waveforms obtained from the Masimo oximeter. The software calculates the initial CRI by comparing the first 30 waveforms after initialization to reference waveforms obtained from humans during progressive central hypovolemia induced by LBNP, and subsequently provides a real-time estimation of physiologic reserve after every heartbeat. Compared with other approaches, a major added benefit of the CRI machine-learning technology is the introduction of a capability to provide high-speed computer techniques that allow for the processing of 100 million data points per second to assess 200 specific features of each continuous noninvasive PPG waveform. This feature of the machine learning is what provides CRI with such high accuracy, sensitivity, and specificity. By comparing subtle changes in arterial waveforms during progressive central hypovolemia to reference waveforms, the algorithm predicts the proportion of compensatory mechanisms available to compensate for changes in effective circulating blood volume. For clinical simplicity, the CRI has been normalized on a scale of 1 to 0 (100% to 0%), where a value of 1 indicates the maximum capacity of compensatory physiologic mechanisms (such as baroreflexes and vasoconstriction) to offset the effects of declining central blood volume, and a value of 0 indicates imminent hemodynamic decompensation.
Pulse pressure variation (PPV)
Pulse pressure variation is defined as the cyclic change in the photoplethysmographic (PPG) waveform derived from the oximeter infrared light absorption during the respiratory cycle. Data obtained from previous research demonstrated that PPV provided a reliable indicator of fluid responsiveness in patients with hemodynamic compromise (7,18,19) (7,18,19) (7,18,19). As such, its relationship with fluid responsiveness has led to the notion that PPV may be useful as a predictor of changes in intravascular volume during acute blood loss (7).
Peripheral perfusion index (PPI)
The peripheral perfusion index was also derived from the PPG signal obtained from the pulse oximeter. The PPI is calculated as the ratio of the amplitude of the pulsatile component to that of the non-pulsatile component of the light reaching the detector of the pulse oximeter. Because peripheral vasoconstriction will alter the pulsatile component of flow independent of hemoglobin oxygen saturation, this technique has been suggested as a reliable, noninvasive indicator of vasomotor tone (8). Compared with both stroke volume and heart rate, there is preliminary evidence that this measure may be a valuable noninvasive indicator of blood volume status in trauma patients (8).
It is important to note, rather than data generated from an arterial waveform, PPV and PPI data collected in the present investigation are generated from proprietary algorithms developed by Masimo, specifically for application to the waveforms generated by their pulse oximeter. Therefore, we designed our experiments to compare CRI, PPV, and PPI based on data generated from the Masimo pulse oximeter waveforms. If we could obtain Masimo's algorithms for calculations of PPV and PPI on the same waveforms, we could also make the head-to-head comparison on arterial line waveforms. However, this was not possible with the data available.
The primary hypothesis was that compared with both the PPI and PPV, the CRI would have superior discriminative ability to predict the onset of hemodynamic decompensation. In order to conserve the family-wise error rate for these two comparisons (CRI vs. PPI and CRI vs. PPV) at α = 0.05, the two test specific type 1 errors were set at α = 0.025. Generalized estimating equations (GEE) with logit link functions and compound symmetry covariance structures were used to perform repeated measures logistic regression analysis. There were three GEE models performed in this study that regressed the onset of hemodynamic decompensation on each of the three measures of interest (i.e., PPI, PPV, and CRI). Based on predicted probabilities from these three GEE models, receiver operating characteristic area under the curve (ROCAUC), sensitivity, and specificity were used to evaluate the ability of all three measures to predict hemodynamic decompensation at each time point in the progressive LBNP-simulated hemorrhage experiment. Last, the predicted probabilities of the onset of hemodynamic decompensation for all three measures were reported by LBNP stepwise chamber decompression levels (baseline to −80 mmHg).
There were two secondary hypotheses for this study. First, compared with CRI scores at the previous LBNP level, CRI scores at the next higher level of progressive simulated hemorrhage would be significantly lower. For example, CRI scores at baseline would be higher than CRI scores at −15 mmHg, CRI scores at −15mmHg would be higher than CRI scores at −30 mmHg, and so forth. Second, compared with low tolerant participants, high tolerant participants would have significantly higher CRI scores at −15 mmHg, −30 mmHg, and −45 mmHg. In order to account for correlations between repeated measurements, generalized linear mixed models (GLMM) were performed. The assumption of normally distributed random effects was assessed with diagnostic plots of conditional residuals; no violations of this assumption were observed. Regarding the first secondary hypothesis, for each of the vital signs of interest, both least squares (LS) means and 95% confidence intervals were reported by LBNP stepwise chamber decompression levels (baseline to −80 mmHg). Regarding the other secondary hypothesis, for each of the vital signs of interest, LS means were reported by stratified tolerance level (HT vs. LT) at each LBNP stepwise chamber decompression level (baseline to −45 mmHg). To account for multiple comparisons of tests of LS means differences, the type 1 errors were set at α = 0.002. One participant experiencing hemodynamic compensation at −90 mmHg was not reported in these figures. All analyses were performed using SAS v9.2 (Cary, NC).
Comparisons of AUC values from ROC analysis indicate that CRI (0.90) was superior to PPV (0.79) and PPI (0.56) in discriminative ability to predict hemodynamic decompensation (Table 1 and Fig. 1; both P ≤ 0.0001). Similarly, the CRI (86.3, 77.7) exhibited greater sensitivity and specificity than both PPV (78.4, 69.0) and PPI (70.6, 28.9). The predicted probability of hemodynamic decompensation using the PPI was approximately 13% and remained stable from baseline to −80 mmHg (Fig. 2A). The predicted probability of hemodynamic decompensation based on PPV (Fig. 2B) was stable during the early stepwise chamber decompression levels (baseline to −45 mmHg), ranging from 13% to 16%, yet increased during the later decompression levels (−60 mmHg to −80 mmHg). Lastly, the predicted probability of hemodynamic decompensation using the CRI (Fig. 2C) was less than 12% during the early stepwise chamber decompression levels (baseline to −30 mmHg), and then steadily increased from 28% to 73% during the remainder of the progressive simulated hemorrhage.
Similar to Figure 2, compared with the CRI, sequential measures of PPI and PPV showed less change over successive levels of simulated hemorrhage. For example, the LS means PPI value remained relatively stable (Fig. 3A). The only statistically significant LS means difference was comparing values at baseline to −15 mmHg (2.44 vs. 1.52; P < 0.0001). The same was true for PPV during the early stepwise chamber decompression levels (Fig. 3B). Comparing baseline vs. −15 mmHg (8.18 vs. 8.17; P = 0.99) and −15 mmHg vs. −30 mmHg (8.17 vs. 8.51; P = 0.57), there was no statistically significant LS means difference. However, during the later stages of simulated hemorrhage (−45 mmHg to −80 mmHg), all sequential comparisons were statistically significant (P < 0.002). In contrast to PPI and PPV, the LS means CRI decreased steadily from baseline to −80 mmHg (Fig. 3C). With the exception of the −70 mmHg vs. −80 mmHg comparison, all other sequential comparisons (baseline to −70 mmHg) were statistically significant (P < 0.002).
Comparison of LS means for each measure stratified by HT vs. LT revealed nearly identical PPI and PPV values for both groups (Fig. 4A and B). For all sequential comparisons of both measures (baseline to −45 mmHg), there were no statistically significant differences between HT and LT participants (all P > 0.002). Similarly, HT participants and LT participants began with similar CRI values at baseline (0.85 vs. 0.78; P = 0.09). However, CRI scores quickly separated by tolerance level and LS means remained higher for HT versus LT participants (Fig. 4C). For example, HT participants had higher CRI scores at all lower levels of simulated hemorrhage: −15 mmHg (0.74 vs. 0.62; P = 0.002); −30 mmHg (0.59 vs. 0.42; P = 0.0001); and −45 mmHg (0.40 vs. 0.22; P < 0.0001).
In comparison to PPV and PPI, results from our study support the primary hypothesis that CRI provides increased sensitivity and specificity as a predictor of hemodynamic decompensation in a human model of progressive hemorrhage. These results support the notion that CRI is a triage decision-support tool capable of providing an earlier warning of progression toward the onset of shock for individual patients compared with either PPV or PPI. This greater specificity of CRI most likely is due to its unique capability to provide an accurate distinction between individuals with low and high tolerance to progressive reduction in central blood volume, assuring those patients with greater risk for developing shock will be appropriately triaged.
Traditionally, standard approaches in the evaluation of trauma patients have included static or “snapshot” measures of cardiovascular status such as heart rate and blood pressure (20). These variables often show little change from population-based normalized averages until the patient is close to the point of hemodynamic decompensation or circulatory shock (21,22) (21,22). As such, accurate diagnosis of individuals at risk for the development of circulatory shock in advance of changes in standard vital signs is critical, as early intervention has been shown to lower morbidity and mortality (23,24) (23,24), specifically in military battlefield casualties (25–29) (25–29) (25–29) (25–29) (25–29). Recent developments in both monitoring capabilities and analysis of hemodynamic waveforms recorded by standard medical monitors such as ECG and pulse oximetry now enable dynamic, real-time analysis of cardiovascular status on a beat-to-beat basis (30). Further, these technological advances allowed more sensitive and specific indicators of cardiovascular status such as the CRI to be developed and compared to traditional vital signs (17).
The results of an early study of 15 mechanically ventilated patients undergoing controlled blood loss suggested that variability in systolic blood pressure was a better predictor of cardiac output changes in response to blood volume loss than were central filling pressures (7). Similarly, compared with traditional vital signs, results from three recent blood loss studies show that CRI has better discriminative ability to predict the onset of hemodynamic decompensation for both low (11,31) (11,31) and high volume blood (9) loss. Our results are consistent with these previous findings, and provide additional evidence that CRI offers a more sensitive and specific method for tracking compensatory status during relative blood volume deficit than other measures.
Traditional approaches used for assessing the status of patients with hemorrhage and shock have been limited by statistical analyses that provide population averages without consideration of significant inter- and intra-individual variance. The result of such an approach is the failure to identify important differences in individual compensatory capacities. It is well documented in the literature that low tolerance to reduced central blood volume in humans is associated with blunted compensatory “reserves” such as an inability to mount robust elevations in heart rate, vasoconstriction, sympathetic nerve activity, and vasoconstrictor hormones (32,33) (32,33). The clinical literature hints at such a cohort of trauma patients with attenuated compensatory responses that can be associated with increased mortality (34,35) (34,35). As a result, more attention is being given to the need for management of shock based on individualized hemodynamic diagnosis and therapy (36). Unlike PPI, PPV, and other standard vital signs, the algorithm for determining CRI is unique in that it captures the status of individuals by estimating the relative amount of reserve remaining for compensation during progressive reductions in central blood volume (10). In contrast to PPI and PPV, it is noteworthy that CRI measured in the present investigation was able to provide an earlier indication of reduced central blood volume. In addition, the CRI measured consistently lower scores for individuals with low tolerance (low compensatory reserve) from those with high tolerance to central hypovolemia. For example, based on PPV, the predicted probability of the onset of hemodynamic decompensation only increased during the later stages of high level hemorrhage and the scores were nearly identical for high versus low tolerant participants. In contrast, based on the CRI, the predicted probability of the onset of hemodynamic decompensation consistently increased from baseline to later stages of high level hemorrhage. In addition, compared with high tolerant participants, CRI scores for low tolerant participants dropped precipitously and remained consistently lower suggesting a faster progression to decompensation. Whether compared with traditional vital signs or newer metrics such as the PPI and PPV, the higher sensitivity and specificity of the CRI may be due to its unique capability for earlier detection of progressive hemorrhage (11,31) (11,31) and ability to differentiate participants with relatively high (large compensatory reserve) from low (small compensatory reserve) tolerance to blood loss (10).
Measurement of PPI during progressive reduction in central blood volume proved to have a relatively low predictive power as indicated by an ROCAUC of only 0.56. Perfusion of peripheral tissue is dependent on the compensatory response of arterial vasoconstriction. Relatively low specificities for changes in peripheral vascular resistance (vascular resistance) and cardiac output during actual blood loss in humans suggest that there exist individual strategies for compensation to central hypovolemia (i.e., some people rely primarily on increases in cardiac output while others rely primarily on peripheral vascular changes) (9). Given the low specificity of systemic vascular resistance as a predictor of blood loss (9), the limited predictive power of PPI as a single indicator of compensation is not unexpected. In contrast, the high predictive power of CRI represents the importance of measuring features of the entire waveform which represents the integration of cardiac mechanisms (ejection wave) and peripheral vascular changes (reflective wave).
Although not as predictive as CRI, PPV displayed significantly greater predictive power than PPI or other vital signs measured in previous experiments (9). The relatively high ROCAUC observed for PPV could be expected since its measurement is based on some features of the arterial waveform. The increased predictive power of CRI compared with PPV probably reflects the higher sensitivity of a machine-learning algorithm designed to analyze 200 features on each waveform. It is this capability that underscores the consistent observation that CRI provides greater sensitivity and specificity than PPV, PPI, cardiac output, peripheral vascular resistance, and other standard vital sign and hemodynamic measurements.
The primary limitation of this study was that it was a simulated hemorrhage experiment among healthy participants and not actual trauma patients. However, regarding the fact that this study used a simulated hemorrhage experiment, the sensitivity and specificity of the CRI reported from this experiment is remarkably similar to results of a voluntary hemorrhage (9,11,31) (9,11,31) (9,11,31). The primary strength of this study is that the CRI is a noninvasive metric that has superior sensitivity and specificity in predicting hemodynamic decompensation. These results support a growing body of literature, suggesting that the CRI has promise as both a clinically meaningful predictor of hemorrhage and in optimizing resuscitation due to circulatory shock (9).
An important factor of any experimental data is its potential translation into clinical application in unhealthy subjects. In actual bleeding, mediators can be released that can alter the dilatory status of the peripheral vascular bed. We hypothesize that such alterations in status of the peripheral vascular bed should be accurately assessed by the algorithm in time interval and feature changes of the reflective wave. To test this hypothesis, there are several ongoing investigations in which CRI data are being collected on patients with various conditions of reduced circulating central blood volume that include Dengue hemorrhagic fever, renal dialysis, orthostatic intolerance, child birth, burn injuries, sepsis, CPR, and other clinical ailments (unpublished data). Preliminary data collected on trauma patients have corroborated the usefulness of CRI for clinical translation. In this study (37), the clinical status of 30 patients with blunt trauma and negligible bleeding was successfully distinguished from 12 patients with penetrating trauma and severe hemorrhage using the CRI. Compensatory reserve values sampled before fluid resuscitation led to a classification accuracy between bleeding and non-bleeding patients of 93%, with a sensitivity of 0.933, specificity of 0.917, and area under the receiver operating characteristic curve of 0.975. For bleeding patients, compensatory reserve increased after receiving IV fluids and subsequently decreased, suggesting continued blood loss and/or waning of compensatory mechanisms. In contrast, non-bleeding patients showed no change or an increase in compensatory reserve with IV fluids.
Compared with both PPI and PPV, the CRI increases the early and accurate prediction of the onset of hemodynamic decompensation. Opportunities to assess the responses of patients with various clinical conditions of inadequate tissue perfusion (e.g., hemorrhage, orthostatic compromise, severe dehydration) that challenge the body's reserve to compensate will allow continued validation of CRI as a decision-support tool that can provide a more sensitive and specific triage tool for early prediction and appropriate timing of interventions to prevent circulatory shock.
1. Eastridge BJ, Mabry RL, Seguin P, Cantrell J, Tops T, Uribe P, Mallett O, Zubko T, Oetjen-Gerdes L, Rasmussen TE, et al. Death on the battlefield (2001-2011): Implications for the future of combat casualty care. J Trauma Acute Care Surg
2012; 73 (6 Suppl 5):S431–S437.
2. Evans JA, van Wessem KJ, McDougall D, Lee KA, Lyons T, Balogh ZJ. Epidemiology of traumatic deaths: Comprehensive population-based assessment. World J Surg
2010; 34 1:158–163.
3. Hinojosa-Laborde C, Rickards CA, Ryan KL, Convertino VA. Heart rate variability during simulated hemorrhage
with lower body negative pressure in high and low tolerant subjects. Front Physiol
4. Parks JK, Elliott AC, Gentilello LM, Shafi S. Systemic hypotension is a late marker of shock
after trauma: A validation study of advanced trauma life support principles in a large national sample. Am J Surg
2006; 192 6:727–731.
5. Patel N, Branson R, Salter M, Henkel S, Seeton R, Khan M, Solanki D, Koutrouvelis A, Li H, Indrikovs A, et al. 2014 military supplement” intrathoracic pressure regulation augments stroke volume and ventricular function in human hemorrhage
2015; 44 (1 Suppl):55–62.
6. Convertino VA, Ryan KL, Rickards CA, Salinas J, McManus JG, Cooke WH, Holcomb JB. Physiological and medical monitoring for en route care of combat casualties. J Trauma
2008; 64 (4 Suppl):S342–353.
7. Ornstein E, Eidelman LA, Drenger B, Elami A, Pizov R. Systolic pressure variation predicts the response to acute blood loss. J Clin Anesth
1998; 10 2:137–140.
8. van Genderen ME, Bartels SA, Lima A, Bezemer R, Ince C, Bakker J, van Bommel J. Peripheral perfusion index as an early predictor for central hypovolemia
in awake healthy volunteers. Anesth Analg
2013; 116 2:351–356.
9. Convertino V, Howard J, Hinojosa-Laborde C, Cardin S, Batchelder P, Mulligan J, Grudic G, Moulton S, MacLeod D. Individual-specific, beat-to-beat trending of significant human blood loss: The compensatory reserve. Shock
2015; 44 (1 Suppl):27–32.
10. Convertino VA, Grudic G, Mulligan J, Moulton S. Estimation of individual-specific progression to impending cardiovascular instability using arterial waveforms. J Appl Physiol (1985)
2013; 115 8:1196–1202.
11. Nadler R, Convertino VA, Gendler S, Lending G, Lipsky AM, Cardin S, Lowenthal A, Glassberg E. The value of noninvasive measurement of the compensatory reserve index in monitoring and triage of patients experiencing minimal blood loss. Shock
2014; 42 2:93–98.
12. Hinojosa-Laborde C, Shade RE, Muniz GW, Bauer C, Goei KA, Pidcoke HF, Chung KK, Cap AP, Convertino VA. Validation of lower body negative pressure as an experimental model of hemorrhage
. J Appl Physiol (1985)
2014; 116 4:406–415.
13. Cooke WH, Ryan KL, Convertino VA. Lower body negative pressure as a model to study progression to acute hemorrhagic shock
in humans. J Appl Physiol (1985)
2004; 96 4:1249–1261.
14. Convertino VA. Endurance exercise training: Conditions of enhanced hemodynamic responses and tolerance to lbnp. Med Sci Sports Exerc
1993; 25 6:705–712.
15. Rickards CA, Ryan KL, Cooke WH, Convertino VA. Tolerance to central hypovolemia
: The influence of oscillations in arterial pressure and cerebral blood velocity. J Appl Physiol (1985)
2011; 111 4:1048–1058.
16. Sather TM, Goldwater DJ, Montgomery LD, Convertino VA. Cardiovascular dynamics associated with tolerance to lower body negative pressure. Aviat Space Environ Med
1986; 57 5:413–419.
17. Moulton SL, Mulligan J, Grudic GZ, Convertino VA. Running on empty? The compensatory reserve index. J Trauma Acute Care Surg
2013; 75 6:1053–1059.
18. Michard F. Changes in arterial pressure during mechanical ventilation. Anesthesiology
2005; 103 2:419–428.quiz 449-415.
19. Michard F, Lopes MR, Auler JO Jr. Pulse pressure variation: Beyond the fluid management of patients with shock
. Crit Care
2007; 11 3:131.
20. Convertino VA, Moulton SL, Grudic GZ, Rickards CA, Hinojosa-Laborde C, Gerhardt RT, Blackbourne LH, Ryan KL. Use of advanced machine-learning techniques for noninvasive monitoring of hemorrhage
. J Trauma
2011; 71 (1 Suppl):S25–S32.
21. Cooke WH, Salinas J, Convertino VA, Ludwig DA, Hinds D, Duke JH, Moore FA, Holcomb JB. Heart rate variability and its association with mortality in prehospital trauma patients. J Trauma
2006; 60 2:363–370.discussion 370.
22. Cooke WH, Salinas J, McManus JG, Ryan KL, Rickards CA, Holcomb JB, Convertino VA. Heart period variability in trauma patients may predict mortality and allow remote triage. Aviat Space Environ Med
2006; 77 11:1107–1112.
23. Pinsky MR. Hemodynamic evaluation and monitoring in the icu. Chest
2007; 132 6:2020–2029.
24. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M. Early goal-directed therapy in the treatment of severe sepsis and septic shock
. N Engl J Med
2001; 345 19:1368–1377.
25. Cap AP, Spinella PC, Borgman MA, Blackbourne LH, Perkins JG. Timing and location of blood product transfusion and outcomes in massively transfused combat casualties. J Trauma Acute Care Surg
2012; 73 (2 Suppl 1):S89–S94.
26. Chung KK, Juncos LA, Wolf SE, Mann EE, Renz EM, White CE, Barillo DJ, Clark RA, Jones JA, Edgecombe HP, et al. Continuous renal replacement therapy improves survival in severely burned military casualties with acute kidney injury. J Trauma
2008; 64 (2 Suppl):S179–S185.discussion S185-177.
27. Gerhardt RT, Berry JA, Blackbourne LH. Analysis of life-saving interventions performed by out-of-hospital combat medical personnel. J Trauma
2011; 71 (1 Suppl):S109–113.
28. Kragh JF Jr, Walters TJ, Baer DG, Fox CJ, Wade CE, Salinas J, Holcomb JB. Survival with emergency tourniquet use to stop bleeding in major limb trauma. Ann Surg
2009; 249 1:1–7.
29. Spinella PC, Perkins JG, Grathwohl KW, Beekley AC, Holcomb JB. Warm fresh whole blood is independently associated with improved survival for patients with combat-related traumatic injuries. J Trauma
2009; 66 (4 Suppl):S69–S76.
30. Convertino VA. Blood pressure measurement for accurate assessment of patient status in emergency medical settings. Aviat Space Environ Med
2012; 83 6:614–619.
31. Stewart C, Mulligan J, Grudic G, Convertino V, Moulton S. Detection of low-volume blood loss: The compensatory reserve index versus traditional vital signs. Journal of Trauma and Acute Care Surgery
2014; 77 6:892–897.
32. Convertino VA. Neurohumoral mechanisms associated with orthostasis: Reaffirmation of the significant contribution of the heart rate response. Front Physiol
33. Convertino VA, Rickards CA, Ryan KL. Autonomic mechanisms associated with heart rate and vasoconstrictor reserves. Clin Auton Res
2012; 22 3:123–130.
34. Vayer JS, Henderson JV, Bellamy RF, Galper AR. Absence of a tachycardic response to shock
in penetrating intraperitoneal injury. Ann Emerg Med
1988; 17 3:227–231.
35. Victorino GP, Battistella FD, Wisner DH. Does tachycardia correlate with hypotension after trauma? J Am Coll Surg
2003; 196 5:679–684.
36. Saugel B, Trepte CJ, Heckel K, Wagner JY, Reuter DA. Hemodynamic management of septic shock
: Is it time for “individualized goal-directed hemodynamic therapy” and for specifically targeting the microcirculation? Shock
2015; 43 6:522–529.
37. Stewart C, Mulligan J, Grudic G, Talley M, Jurkovich G, Convertino V, Moulton S. Noninvasive detection of acute volume loss and fluid resuscitation in trauma patients: Results of a prospective clinical trial. Shock
2015; 43 6:72–73.