External validation of a smartphone app model to predict the need for massive transfusion using five different definitions.

BACKGROUND: Previously, a model to predict massive transfusion protocol (MTP) (activation) was derived using a single-institution data set. The PRospective, Observational, Multicenter, Major Trauma Transfusion database was used to externally validate this model's ability to predict both MTP activation and massive transfusion (MT) administration using multiple MT definitions. METHODS: The app model was used to calculate the predicted probability of MTP activation or MT delivery. The five definitions of MT used were: (1) 10 units packed red blood cells (PRBCs) in 24 hours, (2) Resuscitation Intensity score ≥ 4, (3) critical administration threshold, (4) 4 units PRBCs in 4 hours; and (5) 6 units PRBCs in 6 hours. Receiver operating curves were plotted to compare the predicted probability of MT with observed outcomes. RESULTS: Of 1,245 patients in the data set, 297 (24%) met definition 1, 570 (47%) met definition 2, 364 (33%) met definition 3, 599 met definition 4 (49.1%), and 395 met definition 5 (32.4%). Regardless of the outcome (MTP activation or MT administration), the predictive ability of the app model was consistent: when predicting activation of the MTP, the area under the curve for the model was 0.694 and when predicting MT administration, the area under the curve ranged from 0.695 to 0.711. CONCLUSION: Regardless of the definition of MT used, the app model demonstrates moderate ability to predict the need for MT in an external, homogenous population. Importantly, the app allows the model to be iteratively recalibrated ("machine learning") and thus could improve its predictive capability as additional data are accrued. LEVEL OF EVIDENCE: Diagnostic test study/Prognostic study, level III.

Increased platelet:RBC ratios are associated with improved survival after massive transfusion.

BACKGROUND: Several recent military and civilian trauma studies demonstrate that improved outcomes are associated with early and increased use of plasma-based resuscitation strategies. However, outcomes associated with platelet transfusions are poorly characterized. We hypothesized that increased platelet:red blood cells (RBC) ratios would decrease hemorrhagic death and improve survival after massive transfusion (MT). METHODS: A transfusion database of patients transported from the scene to 22 Level I Trauma Centers over 12 months in 2005 to 2006 was reviewed. MT was defined as receiving ≥ 10 RBC units within 24 hours of admission. To mitigate survival bias, 25 patients who died within 60 minutes of arrival were excluded from analysis. Six random donor platelet units were considered equal to a single apheresis platelet unit. Admission and outcome data associated with the low (>1:20), medium (1:2), and high (1:1) platelet:RBC ratios were examined. These groups were based on the median value of the tertiles for the ratio of platelets:RBC units. RESULTS: Two thousand three hundred twelve patients received at least one unit of blood and 643 received an MT. Admission vital signs, INR, temperature, pH, Glasgow Coma Scale, Injury Severity Score, and age were similar between platelet ratio groups. The average admission platelet counts were lower in the patients who received the high platelet:RBC ratio versus the low ratio (192 vs. 216, p = 0.03). Patients who received MT were severely injured, with a mean (± standard deviation) Injury Severity Score of 33 ± 16 and received 22 ± 15 RBCs and 11 ± 14 platelets within 24 hours of injury. Increased platelet ratios were associated with improved survival at 24 hours and 30 days (p < 0.001 for both). Truncal hemorrhage as a cause of death was decreased (low: 67%, medium: 60%, high: 47%, p = 0.04). Multiple organ failure mortality was increased (low: 7%, medium: 16%, high: 27%, p = 0.003), but overall 30-day survival was improved (low: 52%, medium: 57%, high: 70%) in the high ratio group (medium vs. high: p = 0.008; low vs. high: p = 0.007). CONCLUSION: Similar to recently published military data, transfusion of platelet:RBC ratios of 1:1 was associated with improved early and late survival, decreased hemorrhagic death and a concomitant increase in multiple organ failure-related mortality. Based on this large retrospective study, increased and early use of platelets may be justified, pending the results of prospective randomized transfusion data.

High ratios of plasma and platelets to packed red blood cells do not affect mortality in nonmassively transfused patients.

BACKGROUND: Administration of high transfusion ratios in patients not requiring massive transfusion might be harmful. We aimed to determine the effect of high ratios of fresh frozen plasma (FFP) and platelets (PLT) to packed red blood cells (PRBC) in nonmassively transfused patients. METHODS: Records of 1,788 transfused trauma patients who received <10 units of PRBC in 24 hours at 23 United States Level I trauma centers were reviewed. The relationship between ratio category (low and high) and in-hospital mortality was assessed with propensity-adjusted multivariate proportional hazards models. RESULTS: At baseline, patients transfused with a high FFP:PRBC ratio were younger, had a lower Glasgow Coma Scale score, and a higher Injury Severity Score. Those receiving a high PLT:PRBC ratio were older. The risk of in-hospital mortality did not vary significantly with FFP:PRBC ratio category. Intensive care unit (ICU)-free days, hospital-free days, and ventilator-free days did not vary significantly with FFP:PRBC ratio category. ICU-free days and ventilator-free days were significantly decreased among patients in the high (≥1:1) PLT:PRBC category, and hospital-free days did not vary significantly with PLT:PRBC ratio category. The analysis was repeated using 1:2 as the cutoff for high and low ratios. Using this cutoff, there was still no difference in mortality with either FFP:PRBC ratios or platelet:PRBC ratios. However, patients receiving a >1:2 ratio of FFP:PRBCs or a >1:2 ratio PLT:PRBCs had significantly decreased ICU-free days and ventilator-free days. CONCLUSIONS: FFP:PRBC and PLT:PRBC ratios were not associated with in-hospital mortality. Depending on the threshold analyzed, a high ratio of FFP:PRBC and PLT:PRBC transfusion was associated with fewer ICU-free days and fewer ventilator-free days, suggesting that the damage control infusion of FFP and PLT may cause increased morbidity in nonmassively transfused patients and should be rapidly terminated when it becomes clear that a massive transfusion will not be required.

A normal platelet count may not be enough: the impact of admission platelet count on mortality and transfusion in severely injured trauma patients.

BACKGROUND: Platelets play a central role in hemostasis after trauma. However, the platelet count of most trauma patients does not fall below the normal range (100-450 × 10(9)/L), and as a result, admission platelet count has not been adequately investigated as a predictor of outcome. The purpose of this study was to examine the relationship between admission platelet count and outcomes after trauma. METHODS: A retrospective cohort study of 389 massively transfused trauma patients. Regression methods and the Kruskal-Wallis test were used to test the association between admission platelet count and 24-hour mortality and units of packed red blood cells (PRBCs) transfused. RESULTS: For every 50 × 10(9)/L increase in admission platelet count, the odds of death decreased 17% at 6 hours (p = 0.03; 95% confidence interval [CI], 0.70-0.99) and 14% at 24 hours (p = 0.02; 95% CI, 0.75-0.98). The probability of death at 24 hours decreased with increasing platelet count. For every 50 × 10(9)/L increase in platelet count, patients received 0.7 fewer units of blood within the first 6 hours (p = 0.01; 95% CI, -1.3 to -0.14) and one less unit of blood within the first 24 hours (p = 0.002; 95% CI, -1.6 to -0.36). The mean number of units of PRBCs transfused within the first 6 hours and 24 hours decreased with increasing platelet count. CONCLUSIONS: Admission platelet count was inversely correlated with 24-hour mortality and transfusion of PRBCs. A normal platelet count may be insufficient after severe trauma, and as a result, these patients may benefit from a lower platelet transfusion threshold. Future studies of platelet number and function after injury are needed.

The association of blood component use ratios with the survival of massively transfused trauma patients with and without severe brain injury.

BACKGROUND: The effect of blood component ratios on the survival of patients with traumatic brain injury (TBI) has not been studied. METHODS: A database of patients transfused in the first 24 hours after admission for injury from 22 Level I trauma centers over an 18-month period was queried to find patients who (1) met different definitions of massive transfusion (5 units red blood cell [RBC] in 6 hours vs. 10 units RBC in 24 hours), (2) received high or low ratios of platelets or plasma to RBC units (<1:2 vs. ≥ 1:2), and (3) had severe TBI (head abbreviated injury score ≥ 3) (TBI+). RESULTS: Of 2,312 total patients, 850 patients were transfused with ≥ 5 RBC units in 6 hours and 807 could be classified into TBI+ (n = 281) or TBI- (n = 526). Six hundred forty-three patients were transfused with ≥ 10 RBC units in 24 hours with 622 classified into TBI+ (n = 220) and TBI- (n = 402). For both high-risk populations, a high ratio of platelets:RBCs (not plasma) was independently associated with improved 30-day survival for patients with TBI+ and a high ratio of plasma:RBCs (not platelets) was independently associated with improved 30-day survival in TBI- patients. CONCLUSIONS: High platelet ratio was associated with improved survival in TBI+ patients while a high plasma ratio was associated with improved survival in TBI- patients. Prospective studies of blood product ratios should include TBI in the analysis for determination of optimal use of ratios on outcome in injured patients.

Effect of high product ratio massive transfusion on mortality in blunt and penetrating trauma patients.

BACKGROUND: Recent data suggest that massively transfused patients have lower mortality rates when high ratios (>1:2) of plasma or platelets to red blood cells (RBCs) are used. Blunt and penetrating trauma patients have different injury patterns and may respond differently to resuscitation. This study was performed to determine whether mortality after high product ratio massive transfusion is different in blunt and penetrating trauma patients. METHODS: Patients receiving 10 or more units of RBCs in the first 24 hours after admission to one of 23 Level I trauma centers were analyzed. Baseline physiologic and biochemical data were obtained. Univariate and logistic regression analyses were performed. Adjusted mortality in patients receiving high (≥ 1:2) and low (<1:2) ratios of plasma or platelets to RBCs was calculated for blunt and penetrating trauma patients. RESULTS: The cohort contained 703 patients. Blunt injury patients receiving a high ratio of plasma or platelets to RBCs had lower 24-hour mortality (22% vs. 31% for plasma, p = 0.007; 20% vs. 30% for platelets, p = 0.032), but there was no difference in 30-day mortality (40% vs. 44% for plasma, p = 0.085; 37% vs. 44% for platelets, p = 0.063). Patients with penetrating injuries receiving a high plasma:RBC ratio had lower 24-hour mortality (21% vs. 37%, p = 0.005) and 30-day mortality (29% vs. 45%, p = 0.005). High platelet:RBC ratios did not affect mortality in penetrating patients. CONCLUSION: Use of high plasma:RBC ratios during massive transfusion may benefit penetrating trauma patients to a greater degree than blunt trauma patients. High platelet:RBC ratios did not benefit either group.

A high fresh frozen plasma: packed red blood cell transfusion ratio decreases mortality in all massively transfused trauma patients regardless of admission international normalized ratio.

BACKGROUND: Coagulopathy is present in 25% to 38% of trauma patients on arrival to the hospital, and these patients are four times more likely to die than trauma patients without coagulopathy. Recently, a high ratio of fresh frozen plasma (FFP) to packed red blood cells (PRBCs) has been shown to decrease mortality in massively transfused trauma patients. Therefore, we hypothesized that patients with elevated International Normalized Ratio (INR) on arrival to the hospital may benefit more from transfusion with a high ratio of FFP:PRBC than those with a lower INR. METHODS: Retrospective multicenter cohort study of 437 massively transfused trauma patients was conducted to determine whether the effect of the ratio of FFP:PRBC on death at 24 hours is modified by a patient's admission INR on arrival to the hospital. Contingency tables and logistic regression were used. RESULTS: Trauma patients who arrived to the hospital with an elevated INR had a greater risk of death than those with a lower INR. However, as the ratio of FFP:PRBC transfused increased, mortality decreased similarly between the INR quartiles. CONCLUSIONS: The mortality benefit from a high FFP:PRBC ratio is similar for all massively transfused trauma patients. This is contrary to the current belief that only coagulopathic trauma patients benefit from a high FFP:PRBC ratio. Furthermore, it is unnecessary to determine whether INR is elevated before transfusing a high FFP:PRBC ratio. Future studies are needed to determine the mechanism by which a high FFP:PRBC ratio decreases mortality in all massively transfused trauma patients.

Profoundly abnormal initial physiologic and biochemical data cannot be used to determine futility in massively transfused trauma patients.

BACKGROUND: Improvements in prehospital care and resuscitation have led to increases in the number of severely injured patients who are salvageable. Massive transfusion has been increasingly used. Patients often present with markedly abnormal physiologic and biochemical data. The purpose of this study was to identify objective data that can be used to identify clinical futility in massively transfused trauma patients to allow for early termination of resuscitative efforts. METHODS: A multicenter database was used. Initial physiologic and biochemical data were obtained, and mortality was determined for patients in the 5th and 10th percentiles for each variable. Raw data from the extreme outliers for each variable were also examined to determine whether a point of excessive mortality could be identified. Injury scoring data were also analyzed. A classification tree model was used to look for variable combinations that predict clinical futility. RESULTS: The cohort included 704 patients. Overall mortality was 40.2%. The highest mortality rates were seen in patients in the 10th percentile for lactate (77%) and pH (72%). Survivors at the extreme ends of the distribution curves for each variable were not uncommon. The classification tree analysis failed to identify any biochemical and physiologic variable combination predictive of >90% mortality. Patients older than 65 years with severe head injuries had 100% mortality. CONCLUSION: Consideration should be given to withholding massive transfusion for patients older than 65 years with severe head injuries. Otherwise we did not identify any objective variables that reliably predict clinical futility in individual cases. Significant survival rates can be expected even in patients with profoundly abnormal physiologic and biochemical data.

A predictive model for mortality in massively transfused trauma patients.

BACKGROUND: Improvements in trauma systems and resuscitation have increased survival in severely injured patients. Massive transfusion has been increasingly used in the civilian setting. Objective predictors of mortality have not been well described. This study examined data available in the early postinjury period to identify variables that are predictive of 24-hour- and 30-day mortality in massively transfused trauma patients. METHODS: Massively transfused trauma patients from 23 Level I centers were studied. Variables available on patient arrival that were predictive of mortality at 24 hours were entered into a logistic regression model. A second model was created adding data available 6 hours after injury. A third model evaluated mortality at 30 days. Receiver operating characteristic curves and the Hosmer-Lemeshow test were used to assess model quality. RESULTS: Seven hundred four massively transfused patients were analyzed. The model best able to predict 24-hour mortality included pH, Glasgow Coma Scale score, and heart rate, with an area under the receiver operating characteristic curve (AUROC) of 0.747. Addition of the 6-hour red blood cell requirement increased the AUROC to 0.769. The model best able to predict 30-day mortality included the above variables plus age and Injury Severity Score with an AUROC of 0.828. CONCLUSION: Glasgow Coma Scale score, pH, heart rate, age, Injury Severity Score, and 6-hour red blood cell transfusion requirement independently predict mortality in massively transfused trauma patients. Models incorporating these data have only a modest ability to predict mortality and should not be used to justify withholding massive transfusion in individual cases.

Gender-based differences in mortality in response to high product ratio massive transfusion.

BACKGROUND: Recent data suggest that patients undergoing massive transfusion have lower mortality rates when ratios of plasma and platelets to red blood cells (RBCs) of ≥ 1:2 are used. This has not been examined independently in women and men. A gender dichotomy in outcome after severe injury is known to exist. This study examined gender-related differences in mortality after high product ratio massive transfusion. METHODS: A retrospective study was conducted using a database containing massively transfused trauma patients from 23 Level I trauma centers. Baseline demographic, physiologic, and biochemical data were obtained. Univariate and logistic regression analyses were performed. Adjusted mortality in patients receiving high (≥ 1:2) or low (<1:2) ratios of plasma or platelets to RBCs was compared in women and men independently. RESULTS: Seven hundred four patients were analyzed. In males, mortality was lower for patients receiving a high plasma:RBC ratio at 24 hours (20.6% vs. 33.0% for low ratio, p = 0.005) and at 30 days (34.9% vs. 42.8%, p = 0.032). Males receiving a high platelet:RBC ratio also had lower 24-hour mortality (17.6% vs. 31.5%, p = 0.004) and 30-day mortality (32.1% vs. 42.2%, p = 0.045). Females receiving high ratios of plasma or platelets to RBCs had no improvement in 24-hour mortality (p = 0.119 and 0.329, respectively) or 30-day mortality (p = 0.199 and 0.911, respectively). Use of high product ratio transfusions did not affect 24-hour RBC requirements in males or females. CONCLUSION: Use of high plasma:RBC or platelet:RBC ratios in massive transfusion may benefit men more than women. This may be due to gender-related differences in coagulability. Further study is needed to determine whether separate protocols for women and men should be established.

Crystalloid resuscitation improves survival in trauma patients receiving low ratios of fresh frozen plasma to packed red blood cells.

BACKGROUND: Current trauma resuscitation guidelines recommend giving an initial crystalloid bolus as first line for resuscitation. Recent studies have shown a survival benefit for trauma patients resuscitated with high ratios of fresh frozen plasma (FFP) to packed red blood cells (PRBC). Our aim was to determine whether the volume of crystalloid given during resuscitation correlated with differences in morbidity or mortality based on the ratio of FFP:PRBC given. METHODS: This was a retrospective review of 2,473 transfused trauma patients at 23 Level I trauma centers from July 2005 to October 2007. Patients were separated based on the ratios of FFP:PRBC they received (<1:4, 1:4-1:1, and >1:1) and then analyzed for morbidity and mortality based on whether or not they received at least 1 L crystalloid for each unit of PRBC. Outcomes analyzed were 6-hour, 24-hour, and 30-day survival as well as intensive care unit (ICU)-free days, ventilator-free days, and hospital-free days. RESULTS: Massive transfusion patients who received <1:4 ratios of FFP:PRBC had significantly improved 6-hour, 24-hour, and 30-day mortality and significantly more ventilator-free days if they received at least 1 L of crystalloid for each unit of PRBC. Nonmassive transfusion patients who received <1:4 ratios of FFP:PRBC had significantly improved 6-hour, 24-hour, and 30-day mortality and significantly more ICU-free days, ventilator-free days, and hospital-free days if they received at least 1 L crystalloid for each unit of PRBC. In both massive and nonmassive transfusion groups, the survival benefit and morbidity benefit was progressively less for the 1:4 to 1:1 FFP:PRBC groups and >1:1 FFP:PRBC groups. CONCLUSIONS: If high ratios of FFP:PRBC are unable to be given to trauma patients, resuscitation with at least 1 L of crystalloid per unit of PRBC is associated with improved overall mortality.

Specific abbreviated injury scale values are responsible for the underestimation of mortality in penetrating trauma patients by the injury severity score.

BACKGROUND: The Injury Severity Score (ISS) is widely used as a method for rating severity of injury. The ISS is the sum of the squares of the three worst Abbreviated Injury Scale (AIS) values from three body regions. Patients with penetrating injuries tend to have higher mortality rates for a given ISS than patients with blunt injuries. This is thought to be secondary to the increased prevalence of multiple severe injuries in the same body region in patients with penetrating injuries, which the ISS does not account for. We hypothesized that the mechanism-based difference in mortality could be attributed to certain ISS ranges and specific AIS values by body region. METHODS: Outcome and injury scoring data were obtained from transfused patients admitted to 23 Level I trauma centers. ISS values were grouped into categories, and a logistic regression model was created. Mortality for each ISS category was determined and compared with the ISS 1 to 15 group. An interaction term was added to evaluate the effect of mechanism. Additional logistic regression models were created to examine each AIS category individually. RESULTS: There were 2,292 patients in the cohort. An overall interaction between ISS and mechanism was observed (p = 0.049). Mortality rates between blunt and penetrating patients with an ISS between 25 and 40 were significantly different (23.6 vs. 36.1%; p = 0.022). Within this range, the magnitude of the difference in mortality was far higher for penetrating patients with head injuries (75% vs. 37% for blunt) than truncal injuries (26% vs. 17% for blunt). Penetrating trauma patients with an AIS head of 4 or 5, AIS abdomen of 3, or AIS extremity of 3 all had adjusted mortality rates higher than blunt trauma patients with those values. CONCLUSION: Significant differences in mortality between blunt and penetrating trauma patients exist at certain ISS and AIS category values. The mortality difference is greatest for head injured patients.

Variations between level I trauma centers in 24-hour mortality in severely injured patients requiring a massive transfusion.

BACKGROUND: Significant differences in outcomes have been demonstrated between Level I trauma centers. Usually these differences are ascribed to regional or administrative differences, although the influence of variation in clinical practice is rarely considered. This study was undertaken to determine whether differences in early mortality of patients receiving a massive transfusion (MT, ≥ 10 units pf RBCs within 24 hours of admission) persist after adjustment for patient and transfusion practice differences. We hypothesized differences among centers in 24-hour mortality could predominantly be accounted for by differences in transfusion practices as well as patient characteristics. METHODS: Data were retrospectively collected over a 1-year period from 15 Level I centers on patients receiving an MT. A purposeful variable selection strategy was used to build the final multivariable logistic model to assess differences between centers in 24-hour mortality. Adjusted odds ratios for each center were calculated. RESULTS: : There were 550 patients evaluated, but only 443 patients had complete data for the set of variables included in the final model. Unadjusted mortality varied considerably across centers, ranging from 10% to 75%. Multivariable logistic regression identified injury severity score (ISS), abbreviated injury scale (AIS) of the chest, admission base deficit, admission heart rate, and total units of RBC transfused, as well as ratios of plasma:RBC and platelet:RBC to be associated with 24-hour mortality. After adjusting for severity of injury and transfusion, treatment variables between center differences were no longer significant. CONCLUSIONS: In the defined population of patients receiving an MT, between-center differences in 24-hour mortality may be accounted for by severity of injury as well as transfusion practices.

The thermodilution technique for measuring resting energy expenditure does not agree with indirect calorimetry for the critically ill patient.

BACKGROUND: The complications associated with overfeeding critically ill patients are well documented. Indirect calorimetry is touted as the gold standard for measuring resting energy expenditure (REE). Unfortunately, the device is expensive, and many centers do not have this technology. The thermodilution technique for measuring cardiac output and calculating REE using the Fick equation has been reported to be an acceptable alternative. This study compared these techniques in a critically ill population. METHODS: Forty consecutive patients with indwelling Swan-Ganz catheters in the surgical intensive care unit were prospectively studied while under the consultative care of the nutrition support service. REE was determined in all patients by both techniques within a 2-hour period. An error of 5% (approximately+/-100 kcal/d) between the two methods was deemed acceptable for clinical use. RESULTS: Mean values for REE were 1928+/-558 vs 1898+/-518 kcal/d for the indirect calorimetry and thermodilution methods, respectively, and were not significantly different. However, there was great variation between the two techniques for the majority of patients such that REE determinations did not agree (t = 6.8; p < .0005). In 70% of the patients, REE determinations differed by > or =20% and in 10% of the patients by 50%. Additionally, the greater the difference between the two methods, the more the thermodilution method tended to overestimate REE. CONCLUSIONS: When compared with indirect calorimetry in a critically ill population, the thermodilution method demonstrated an intersubject variability that is unacceptable for clinical use.

Diaphragm-like strictures of the ileum associated with NSAID use: a rare complication.

Nonsteroidal anti-inflammatory drugs (NSAIDs) are widely prescribed for many conditions including arthritis. A rare complication of their use is diaphragm-like strictures of the small and large intestines. A 65-year-old woman with a 12-year history of arthritis came to us with a 35-pound weight loss and anorexia. She had been taking piroxicam for 3 years. Evaluation including enteroclysis revealed multiple mid-ileal diaphragm-like strictures and proximal small bowel dilatation. The symptoms persisted despite discontinuance of the drug. Abdominal exploration with intraoperative enteroscopy revealed five ileal strictures within a short segment of bowel. Resection was done and completion enteroscopy showed no other strictures. The patient recovered uneventfully and had full resolution of the symptoms. We discuss the difficulties in diagnosis and management of this drug complication and briefly review the literature.

Laparoscopic evacuation of a subcapsular renal hematoma causing symptomatic hypertension.

We report the case of a subcapsular hematoma following extracorporeal shockwave lithotripsy which presented as symptomatic hypertension. When medical therapy proved ineffective, laparoscopic decompression of the hematoma corrected the hypertension.