Whats new in critical illness and injury science: Predicting mortality in trauma!

The study “Predictors of one year mortality in adult injured patients admitted to the trauma centre of KGMU” is probably the first study that focuses on 1-year mortality of injured. It is unique because it attempts to obtain estimates of external validity, it uses Cox proportional hazard model to quantify the effect of predictors over a 1-year period and it has been done in a resource-constrained scenario in a country which lacks formal trauma care systems and the findings may be applicable in similar scenarios in developing nations of the world. However, certain important aspects need to be highlighted that may be incorporated in similar studies conducted in future.

Authors have used Injury Severity Score (ISS) to characterize the severity of trauma. ISS is an internationally accepted score to measure injury severity. Since ISS considers only one injury per body region, it is constrained to consider a second, perhaps less severe injury in the second body region than a second, more serious injury in the first injury. The original intent of the ISS algorithm to consider the body as a whole is in conflict with more fundamental principle that more severe injuries should be considered over less severe injuries. An alternative that could have been used by the authors is A Severity Characterization of Trauma (ASCOT), which includes descriptors of anatomical injury, patient physiology on Emergency Department admission, patient age, and type of injury. ASCOT uses the four Anatomical Profile (AP) components (A, B, C, and D) to describe anatomical injury and the three coded values of Revised Trauma Score (RTS) variables to describe physiology. Age groups are less than 55 years, by decade from 55 to 84, and 85 years or older. Patients are separated into blunt and penetrating injury subsets for analysis. For each set, patients with extremely poor or good prognoses are excluded from the logistic function modeling. These include seriously injured patients with an Abbreviated Injury Scale (AIS) 6 injury and/or who arrive at the emergency department (ED) in cardiac arrest and patients with max AIS 1 or 2 and RTS > 0. In ASCOT, the representation of anatomic injury provided by the AP components obviates many short comings of ISS, for example, it maintains separate scores by body region/system and it includes AIS scores for all serious injures with a region, not just for the most severe, and it identifies substantially more homogeneous subgroups for evaluation and prediction than ISS.[1] It therefore may have more face validity than the ISS, and be more useful to clinicians by providing a more interpretable description of injury.[1]

Two important indicators of shock; namely, serum lactate and base deficit were not included in the study due to logistic reasons, which is very reasonable considering the fact that the study was done in a resource-constrained scenario. However, it needs to be pointed out that lactic acidosis may indicate tissue hypoperfusion even when vital signs are initially normal or relatively normal. Numerous studies have established the use of lactate as a diagnostic, therapeutic, and prognostic marker of global tissue hypoxia in circulatory shock.[234567] Previous studies have shown that a lactate concentration >4 mmol/L in the presence of the systemic inflammatory response syndrome (SIRS) criteria, significantly increases intensive care unit admission rates and mortality rate in normotensive patients.[8910] In addition to allowing for identification of shock or hypoperfusion, lactic acid levels can also provide an endpoint of resuscitation and indicator of restored perfusion. Effective lactate clearance has been found to be associated with lower mortality levels in trauma, sepsis, and postcardiac arrest.[111213] Base deficit may provide yet another metabolic end point of resuscitation. A retrospective analysis of prospectively collected data on 100 trauma patients by Kincaid and colleagues[14] revealed that patients with persistently elevated base deficit had higher rates of multiple organ failure and death when compared with those with a low base deficit. The most significant shortcoming of base deficit is that other diseases and physiologic states may alter base deficit without reflecting an alteration of perfusion. However, lactate has been found to be more predictive of morbidity and mortality than was base deficit among their study population.[15]

The study reports a very high predictive value (β co efficient of −3. 82; −ve value meaning lower the Glasgow Coma Scale (GCS) score, higher the chance of mortality) of GCS score at admission in predicting mortality. A very important finding of the study is that GCS score at the time of admission is the only valid predictor of 1-year mortality in the injured after 6th day of admission upto 8th week. However, it has been reported that mortality prediction using GCS may not be very accurate due to presence of paralysis, use of sedating medications, and coexisting injuries.[16] In an organized system of trauma care with prehospital care incorporating advanced life support system protocols, use of paralytics in the field and during transport make the early post-resuscitation GCS unavailable. Thus, such high predictive value of GCS may be limited to areas which lack advanced prehospital care protocols, use of paralytics and endotracheal intubation, and the results may not be replicated in areas with advanced prehospital care facilities.

This study validates previously reported predictors of mortality such as age, ISS, abnormal respiratory rate (>26 or ≤10) at admission, increased APTT, and GCS score at admission. However due to lack of information on GCS score in cervical spine injury patients, it is not possible to comment on this class of patients. A future study focusing on cervical spine injury patients may be planned to identify predictors of mortality in this class of patients. The study reports a positive association between referral and below poverty line status and goes onto provide possible explanations and the need of a future study to identify the reasons for the association. This is a very important question that needs to be answered by a study that focuses on referral patterns and reasons for referral at peripheral hospitals.