The most common cause of the discoloration of non-vital teeth:
A- stains from drinking coffee
B- hemorrhage after trauma *****
C- toxins from chronic illnesses
--------------------------------------
About one third of hospital deaths due to trauma are caused by severe bleeding. The lack of early and appropriate treatment in these patients is an avoidable cause of death. The triage criteria (classification of patients prior to treatment according to their severity) that allow rapid identification of high-risk patients can reduce mortality. Recent data on the significant decrease in mortality in these patients with early administration of tranexamic acid, further underline the importance of timely identification of life-threatening bleeding. However, this early prognosis should be based on variables that can be easily measured shortly after the injury. Several clinical variables related to the physiological response to the decrease in intravascular volume predict the risk of death in patients with trauma and hemorrhage. These include blood pressure, capillary filling time, level of consciousness (Glasgow coma score), heart rate and respiratory rate. Prognostic models that combine these variables are necessary in order to increase prognostic accuracy. An exact and easy-to-use model could help professionals in prehospital triage. When used, diagnostic and therapeutic procedures that can save lives (such as surgery and tranexamic acid) would be applied before.
Prognostic models should be based on contemporary data, as therapeutic practices changed and the age of trauma patients increased in high-income countries. In addition, although the majority of trauma deaths occur in low- and middle-income countries, most prognostic models are based on data from high-income countries.
The objective of this work was to create and validate a prognostic model for premature death in patients with trauma hemorrhage.
Methods:
- Creation of the model:
To create the prognostic model, the authors met with professionals from three contexts: prehospital, battlefield and emergency services. They identified the variables and interactions that they considered important and convenient for these work areas and obtained information on the best way to present the model. They incorporated 20,127 patients from the CRASH-2 study who suffered trauma with significant bleeding or who were at risk of suffering it, within eight hours of the injury. The study was conducted in 274 hospitals in 40 countries. The main evaluation criterion was mortality from all causes.
- Prognostic factors:
Variables were taken from the patient admission forms completed before randomization, to be analyzed as possible prognostic factors. The variables were: the demographic characteristics of the patient (age and sex), the characteristics of the trauma (type of trauma [penetrating or closed and penetrating] and time from trauma to randomization) and physiological variables (Glasgow coma score, systolic pressure, heart rate, respiratory rate and central capillary refill time).
- Multifactor analysis:
Initially all prognostic factors were included in the multifactor logistic regression. A variable was also included for the economic region (low, middle or high income country). Interactions were specifically explored according to age and type of trauma.
A stepwise stepwise approach was used. First, all possible prognostic factors and terms of interaction that were considered acceptable were included. These interactions included all possible prognostic factors with the type of trauma, time since trauma and age. Then the terms for which no strong evidence of an association was found, one at a time, were eliminated, according to the values of P (<0 .05="" a="" all="" finally="" from="" in="" model="" p="" reached="" significant.="" statistically="" terms="" test.="" the="" wald="" was="" were="" which="">
- Evaluation:
The authors assessed the prognostic capacity of the model in relation to calibration and discrimination. The first indicates whether the observed risks are in accordance with the forecasted risks. Discrimination indicates whether low-risk patients can be separated from high-risk patients.
- External validation:
For external validation, the authors used data from the Trauma Audit and Research Network (TARN), which includes 60% of hospitals that receive trauma patients in England and Wales and some hospitals in the rest of Europe. Only patients with estimated bleeding in at least 20% were selected, whom the authors considered comparable to patients in the CRASH-2 study.
--------------------------------------
About one third of hospital deaths due to trauma are caused by severe bleeding. The lack of early and appropriate treatment in these patients is an avoidable cause of death. The triage criteria (classification of patients prior to treatment according to their severity) that allow rapid identification of high-risk patients can reduce mortality. Recent data on the significant decrease in mortality in these patients with early administration of tranexamic acid, further underline the importance of timely identification of life-threatening bleeding. However, this early prognosis should be based on variables that can be easily measured shortly after the injury. Several clinical variables related to the physiological response to the decrease in intravascular volume predict the risk of death in patients with trauma and hemorrhage. These include blood pressure, capillary filling time, level of consciousness (Glasgow coma score), heart rate and respiratory rate. Prognostic models that combine these variables are necessary in order to increase prognostic accuracy. An exact and easy-to-use model could help professionals in prehospital triage. When used, diagnostic and therapeutic procedures that can save lives (such as surgery and tranexamic acid) would be applied before.
Prognostic models should be based on contemporary data, as therapeutic practices changed and the age of trauma patients increased in high-income countries. In addition, although the majority of trauma deaths occur in low- and middle-income countries, most prognostic models are based on data from high-income countries.
The objective of this work was to create and validate a prognostic model for premature death in patients with trauma hemorrhage.
Methods:
- Creation of the model:
To create the prognostic model, the authors met with professionals from three contexts: prehospital, battlefield and emergency services. They identified the variables and interactions that they considered important and convenient for these work areas and obtained information on the best way to present the model. They incorporated 20,127 patients from the CRASH-2 study who suffered trauma with significant bleeding or who were at risk of suffering it, within eight hours of the injury. The study was conducted in 274 hospitals in 40 countries. The main evaluation criterion was mortality from all causes.
- Prognostic factors:
Variables were taken from the patient admission forms completed before randomization, to be analyzed as possible prognostic factors. The variables were: the demographic characteristics of the patient (age and sex), the characteristics of the trauma (type of trauma [penetrating or closed and penetrating] and time from trauma to randomization) and physiological variables (Glasgow coma score, systolic pressure, heart rate, respiratory rate and central capillary refill time).
- Multifactor analysis:
Initially all prognostic factors were included in the multifactor logistic regression. A variable was also included for the economic region (low, middle or high income country). Interactions were specifically explored according to age and type of trauma.
A stepwise stepwise approach was used. First, all possible prognostic factors and terms of interaction that were considered acceptable were included. These interactions included all possible prognostic factors with the type of trauma, time since trauma and age. Then the terms for which no strong evidence of an association was found, one at a time, were eliminated, according to the values of P (<0 .05="" a="" all="" finally="" from="" in="" model="" p="" reached="" significant.="" statistically="" terms="" test.="" the="" wald="" was="" were="" which="">
- Evaluation:
The authors assessed the prognostic capacity of the model in relation to calibration and discrimination. The first indicates whether the observed risks are in accordance with the forecasted risks. Discrimination indicates whether low-risk patients can be separated from high-risk patients.
- External validation:
For external validation, the authors used data from the Trauma Audit and Research Network (TARN), which includes 60% of hospitals that receive trauma patients in England and Wales and some hospitals in the rest of Europe. Only patients with estimated bleeding in at least 20% were selected, whom the authors considered comparable to patients in the CRASH-2 study.
- Simple forecast model:
To easily use the patient's place of care, the authors created a simple prognostic model. It included the strongest prognostic factors with the same quadratic and cubic terms used in the complete model, with adjustment for tranexamic acid.
The prognostic model is presented as a graph that crosses these predicted factors recoded into several classes. To assemble the classes, clinical and statistical criteria were considered. In each box of the graph, the risk was estimated for a person with the values of each prognostic factor at the midpoint of the upper and lower limits for that box. The boxes of the graph were colored in four groups according to the probability of death: <6 21-50="" 6-20="" and=""> 50%.
Results:
In total, 3,076 patients (15%) of the 20127 patients in the CRASH-2 study and 1,765 (12%) of the 14,220 TARN died. Age was associated in a positive and increasing way with the risk of death; systolic pressure, heart rate and respiratory rate showed U-shaped relationships; Glasgow coma score had a negative association with the risk of death. In the CRASH-2 study, age was positively associated with mortality for each of the causes of death mentioned. In the multifactorial analysis, Glasgow coma score, systolic pressure and age were the three strongest prognostic factors. Heart rate, respiratory rate and hours since trauma were associated with mortality and were included in the final model. Patients from low and middle income countries were more likely to die in relation to those from high income countries. Although hair filling time was weakly associated with mortality, it was not included in the prognostic model because in situations with poor visibility, such as on the battlefield, it is difficult to determine. Some evidence of statistical interaction was found between the Glasgow coma score and the type of trauma. The low score was associated with a worse prognosis for closed trauma.
Validation:
The model showed good internal validation and good calibration, except in very high risk patients for whom the model predicted more risk than the real one.
For the external validation the same variables were used, except for the hours since the trauma, since for this variable there were too many patients with incomplete data. The discrimination was good and the calibration was satisfactory.
Presentation of the model:
Entering the value of the prognostic factors shows the expected risk of death at 28 days. For example, a 70-year-old patient from a low-income country with a Glasgow coma score of 14, systolic pressure of 100 mm Hg, heart rate of 110 per minute and respiratory rate of 35 per minute, has 32% of probability of death at 28 days.
A simple prognostic model that can be used at the patient's bedside is also important. This model comprises the three strongest prognostic variables: Glasgow coma score, systolic pressure and age.
Consequences of the study:
The effect of age is important, since in the high-income countries the average age of trauma patients is increasing. This effect of age probably reflects the increased incidence of coexisting diseases, especially cardiovascular diseases. Elderly patients are more likely to suffer from coronary heart disease and decreased oxygen supply due to hemorrhage trauma may increase the risk of myocardial ischemia. Another possible explanation for the increased risk of death from occlusive vascular disease is related to the triggering of the inflammatory process after trauma. The triggered inflammatory response comprises the increase in the plasma concentration of interleukin-1, interleukin-2, tumor necrosis factor-α, interleukin-6, interleukin-12 and interferon-γ. In patients with trauma hemorrhage, plasmin activation also occurs, which is key in the fibrinolytic response in the first hours after trauma.
The authors acknowledge that estimating the risk of death in the traumatized patient with bleeding is a challenge. It is a process that uses not only physiological variables, but complementary tests, as well as the response to treatments. A prognostic model could never replace medical criteria, but it can support it.
Conclusions:
This prognostic model can be used to obtain valid prognoses of mortality in patients with trauma hemorrhage, to assist in triage and possibly to apply before the diagnostic and therapeutic procedures that can save lives. Age is an important prognostic factor, especially in high-income countries where the population suffering from trauma is older.
Labels
Promatric