Leopoldo C. Cancio

Accurate and effective diagnosis of actual injury severity can be problematic in trauma patients. Inherent physiologic compensatory mechanisms may prevent accurate diagnosis and mask true severity in many circumstances. The objective of this project was the development and validation of a multiparameter machine learning algorithm and system capable of predicting the need for life-saving interventions (LSIs) in trauma patients. Statistics based on means, slopes, and maxima of various vital sign measurements corresponding to 79 trauma patient records generated over 110,000 feature sets, which were used to develop, train, and implement the system. Comparisons among several machine learning models proved that a multilayer perceptron would best implement the algorithm in a hybrid system consisting of a machine learning component and basic detection rules. Additionally, 295,994 feature sets from 82 h of trauma patient data showed that the system can obtain 89.8 % accuracy within 5 min of recorded LSIs. Use of machine learning technologies combined with basic detection rules provides a potential approach for accurately assessing the need for LSIs in trauma patients. The performance of this system demonstrates that machine learning technology can be implemented in a real-time fashion and potentially used in a critical care environment.

Introduction: The clinical use of serial quantitative computed tomography (CT) to characterize lung disease and guide the optimization of mechanical ventilation in patients with acute respiratory distress syndrome (ARDS) is limited by the risk of cumulative radiation exposure and by the difficulties and risks related to transferring patients to the CT room. We evaluated the effects of tube current-time product (mAs) variations on quantitative results in healthy lungs and in experimental ARDS in order to support the use of low-dose CT for quantitative analysis. Methods: In 14 sheep chest CT was performed at baseline and after the induction of ARDS via intravenous oleic acid injection. For each CT session, two consecutive scans were obtained applying two different mAs: 60 mAs was paired with 140, 15 or 7.5 mAs. All other CT parameters were kept unaltered (tube voltage 120 kVp, collimation 32 × 0.5 mm, pitch 0.85, matrix 512 × 512, pixel size 0.625 × 0.625 mm). Quantitative results obtained at different mAs were compared via Bland-Altman analysis. Results: Good agreement was observed between 60 mAs and 140 mAs and between 60 mAs and 15 mAs (all biases less than 1%). A further reduction of mAs to 7.5 mAs caused an increase in the bias of poorly aerated and nonaerated tissue (-2.9% and 2.4%, respectively) and determined a significant widening of the limits of agreement for the same compartments (-10.5% to 4.8% for poorly aerated tissue and -5.9% to 10.8% for nonaerated tissue). Estimated mean effective dose at 140, 60, 15 and 7.5 mAs corresponded to 17.8, 7.4, 2.0 and 0.9 mSv, respectively. Image noise of scans performed at 140, 60, 15 and 7.5 mAs corresponded to 10, 16, 38 and 74 Hounsfield units, respectively. Conclusions: A reduction of effective dose up to 70% has been achieved with minimal effects on lung quantitative results. Low-dose computed tomography provides accurate quantitative results and could be used to characterize lung compartment distribution and possibly monitor time-course of ARDS with a lower risk of exposure to ionizing radiation. A further radiation dose reduction is associated with lower accuracy in quantitative results.

Total intravenous anesthesia (TIVA) is frequently used for major operations requiring general anesthesia in critically ill burn patients. We reviewed our experience with this approach. During a 22-month period, 547 major burn surgeries were performed in this center's operating room and were staffed by full-time burn anesthesiologists. The records of all 123 TIVA cases were reviewed; 112 records were complete and were included. For comparison, 75 cases were selected at random from a total of 414 non-TIVA general anesthetics. Some patients had more than one operation during the study: as appropriate for the analysis in question, each operation or each patient was entered as an individual case. For inter-patient analysis, exposure to 1 or more TIVAs was used to categorize a patient as member of the TIVA group. Excision and grafting comprised 78.2% of the operations. 14 TIVA regimens were used, employing combinations of 4 i.v. drugs: ketamine (K, 91 cases); i.v. methadone (M, 62); fentanyl (F, 58); and propofol (P, 21). The most common regimens were KM (34 cases); KF (26); KMF (16); and K alone (8). Doses used often exceeded those used in non-burn patients. TIVA was preferred for those patients who were more critically ill prior to surgery, with a higher ASA score (3.87 vs. 3.11). Consistent with this, inhalation injury (26.7 vs. 1.6%), burn size (TBSA, 36.3 vs. 15.8%), and full-thickness burn size (FULL, 19.8 vs. 6.5%) were higher in TIVA than in non-TIVA patients. Despite this, intraoperative pressor use was as common in TIVA as in non-TIVA cases (23.9 vs. 22.7%). TIVA was used in patients whose inhalation injury rate and TBSA were greater than those of non-TIVA patients. TIVA cases were not associated with increased hemodynamic instability. TIVA is a viable approach to general anesthesia in critically ill burn patients.

The relationship between end-tidal carbon dioxide (EtCO₂) and arterial carbon dioxide (PaCO₂)—if better defined—could facilitate the difficult task of ventilation in prehospital trauma patients. We aimed to study the PaCO₂-EtCO₂ relationship before, during, and after chest trauma, hemorrhage, and resuscitation in swine. Twenty-four swine were intubated, anesthetized, and monitored in an animal intensive care unit during three phases: phase 1 (day 1, healthy animals); phase 2 (day 2, injury), which consisted of blunt chest trauma, hemorrhage, and resuscitation; and phase 3 (day 2, after injury). “Respiratory maneuvers” (changes in respiratory rate and tidal volume [TV], intended to vary the PaCO₂ over a range of 25 to 85 mmHg, were performed during phases 1 and 3. End-tidal CO₂ and PaCO₂ were recorded after each respiratory maneuver and analyzed using linear regression. During phase 1, PaCO₂ and EtCO₂ were strongly correlated (r² = 0.97, P < 0.01). During phase 2, animals developed decreased oxygenation (PaO₂:FiO₂ [fraction of inspired oxygen] ratio 2-EtCO₂ relationship deteriorated (r² = 0.25, P < 0.0001). During phase 3, oxygenation, hemodynamics, and the PaCO₂-EtCO₂ relationship recovered (r² = 0.92, P < 0.01). End-tidal CO₂ closely correlates to PaCO₂ in healthy animals and after injury/resuscitation across a wide range of respiratory rates and tidal volumes. Once oxygenation and hemodynamics are restored, EtCO₂ can be used to predict PaCO₂ following chest trauma/hemorrhage and should be considered for patient monitoring. This work demonstrated that EtCO₂ alone can reliably be used to estimate PaCO₂ in uninjured subjects and in those subjects who have been resuscitated from severe injury. Immediately after blunt chest injury, the correlation between EtCO₂ and PaCO₂ is temporarily unstable. Under these circumstances (with abnormal oxygenation and/or hemodynamics), greater caution and other monitoring tools may be required.

Introduction: Acute kidney injury (AKI) is a common and devastating complication in critically ill burn patients with mortality reported to be between 80 and 100%. We aimed to determine the effect on mortality of early application of continuous venovenous hemofiltration (CVVH) in severely burned patients with AKI admitted to our burn intensive care unit (BICU). Methods: We performed a retrospective cohort study comparing a population of patients managed with early and aggressive CVVH compared with historical controls managed conservatively before the availability of CVVH. Patients with total body surface area (TBSA) burns of more than 40% and AKI were treated with early CVVH and their outcomes compared with a group of historical controls. Results: Overall, the 28-day mortality was significantly lower in the CVVH arm (n = 29) compared with controls (n = 28) (38% vs. 71%, P = 0.011) as was the in-hospital mortality (62% vs. 86%, P = 0.04). In a subgroup of patients in shock, a dramatic reduction in the pressor requirement was seen after 24 and 48 hours of treatment. Compared with controls (n = 19), significantly fewer patients in the CVVH group (n = 21) required vasopressors at 24 hours (100% vs 43%, P < 0.0001) and at 48 hours (94% vs 24%, P < 0.0001). In those with acute lung injury (ALI)/acute respiratory distress syndrome (ARDS), there was a significant increase from baseline in the partial pressure of arterial oxygen (PaO₂) to fraction of inspired oxygen (FiO₂) ratio at 24 hours in the CVVH group (n = 16, 174 ± 78 to 327 ± 122, P = 0.003) but not the control group (n = 20, 186 ± 64 to 207 ± 131, P = 0.98). Conclusions: The application of CVVH in adult patients with severe burns and AKI was associated with a decrease in 28-day and hospital mortality when compared with a historical control group, which largely did not receive any form of renal replacement. Clinical improvements were realized in the subgroups of patients with shock and ALI/ARDS. A randomized controlled trial comparing early CVVH to standard care in this high-risk population is planned.

Background: Our purpose was to compare the Revised Trauma Score (RTS) with the new Field Triage Score (FTS) for prediction of mortality (MORT) and of need for massive transfusion (MASS, ≥10 units of packed cells or whole blood) in casualties arriving at combat support hospitals in Iraq. Methods: Six hundred ninety-two cases were reviewed; 536 had complete data and were included. Total Glasgow Coma Scale score (GCS_total) not GCS_motor was used. Thus, a modification (FTS₀₇) of the FTS was calculated, using GCS p = e^k/(1 + e^k). Results: By LRA for MORT, k = 0.616 – 0.438 × RTS; AUC = 0.708. When used instead of RTS, FTS₀₇ provided k = −0.716 – 1.009 × FTS₀₇; AUC = 0.687 (NS). For MASS, k = 0.638 – 0.115 × RTS – 0.011 × DAP + 0.358 × SI, where DAP is diastolic arterial pressure and SI is shock index, i.e., heart rate or SAP; AUC = 0.638. When used instead of RTS, FTS₀₇ provided k = −0.740 – 0.376 × FTS₀₇– 0.011 × DAP; AUC = 0.618 (NS). Conclusions: RTS emerged as the best predictor of MORT, with FTS₀₇ a close surrogate. This indicates the effect of impaired mentation on MORT in these data. For prediction of MASS, RTS as well as the heart rate and blood pressure predominated. The advantage of FTS₀₇ (or original FTS) over RTS is the former’s ease of computation.