Is CXR the “B” of ABCs?

4 comments for “Is CXR the “B” of ABCs?”

1. 

   jkhadpe

   October 1, 2013 at 3:09 pm

   Reposted to hopefully get some more exposure for the article at the top of the blog.

   I just wanted to start off with a couple general comments. I think the overall design of the study is good and I don’t find much fault in their analysis. There’s always room for criticism (like did they routinely perform EFAST exams?), but for the most part it seems well done.

   My main issue with the article is mostly with the premise which is I would not equate a CDR to r/o cspine fx or knee fx with a CDR for blunt thoracic injury. As listed in the article, blunt thoracic injury encompasses a multitude of individual clinically important injuries involving bone, lung, vessels, and not to mention the heart which they didn’t even look at. This is similar logic as to why I think so far all the syncope rules have failed as they are trying to apply a one-size-fits-all rule to what is really multiple possible disease states. Maybe it’s a San Francisco thing. The cspine and ottawa knee/ankle/foot rules work because they are basically looking to rule out one thing – fx.

   Last comment for now, no one will use a CDR that is more than 5 items long.

   One take home point is that the H+P is still king and we don’t need to blindly CT every trauma patient that walks in the door. All the criteria in this rule and similar head CT and cspine CT rules use items from the H+P. We should use our clinical judgement to decide what imaging studies are necessary in trauma patients.

   Curious if others agree or disagree? Also feel free to comment on other aspects of the article.

   • 

     TSmith

     October 4, 2013 at 6:52 am

     Sad, we didn’t get to discuss this at conference. I agree with Jay as far CDR are concerned. There are too many variables to simplify thoracic trauma by CDR. However, the study design of this research was done very well in my opinion. Great efforts were made to eliminate various biases. Given County sees more penetrating trauma; I question how much this CDR would even apply to our patient population.

2. 

   eli.brown

   October 4, 2013 at 12:00 pm

   Thanks for the discussion.

   In brief, this was a prospective multicenter cohort validation study performed at 9 level I trauma centers in the U.S. from Dec 2009 to Jan 2012. The inclusion criteria were: (1) age greater than 14 years, (2) blunt trauma occurring within 24 hours of ED presentation, (3) receiving either CXR or chest CT in the ED as part of the blunt trauma evaluation. Treating physicians were asked to fill out a 1-page survey prior to viewing the chest images indicating the presence of NEXUS Chest DI criteria. This decision instrument had previously been derived through an observational study by Rodriguez et al. to include: (1) age greater than 60 years, (2) rapid deceleration mechanism, (3) chest pain, (4) intoxication, (5) abnormal alertness/mental status, (6) distracting painful injury, and (7) tenderness to chest wall palpation. Radiologic interpretations were performed by board-certified radiologists who were blind to patient enrollment to determine the presence of thoracic injury seen on chest imaging (TICI). To address the clinical significance of certain missed injuries, a panel of 10 trauma surgeons and emergency physicians convened to classify thoracic injuries (none, minor, and major) according to the associated clinical interventions.
   Of the 9905 patients enrolled, 1479 (14.9%) had TICI. Of these, 363 were of major clinical significance, 1079 minor, and 36 of no clinical significance. Regardless of clinical significance, the Nexus Chest DI had: (1) a sensitivity of 98.8% (95% CI, 98.1% – 99.3%), (2) a NPV of 98.5% (95% CI, 97.6% – 99.1%), and (3) a specificity of 13.3% (95% CI, 12.6% – 14.1%). In total, there were 17 false-negative TICI patients, one of major clinical significance (chest tube placement for hemopneumothorax), 14 of minor clinical significance, and 2 of no clinical significance. Of note, 13 of 17 missed TICI occurred at a single site.

3. 

   eli.brown

   October 4, 2013 at 12:41 pm

   So, here are a couple questions to ask yourself….

   Is 7 items too long for a CDR? What if 3 criteria are identical to NEXUS cervical spine criteria?

   Is the fact that 13 of 17 missed thoracic injuries occurred at a single site concerning or reassuring? Is this decision rule generalizable to other institutions, especially non-academic, non-level I trauma centers? Another study by Sears et al. reported that senior trauma surgeon judgement alone had a 95.1% sensitivity for determining the need for chest imaging. This study did not compare judgement alone to the CDI, and it did not demonstrate its effectiveness at non-level I trauma centers.

   Is it a concern that utilizing this decision instrument might increase the amount of chest imaging? I know it’s hard to imagine, but should this really be used for all blunt trauma? What if it’s really low risk? Does everyone over 60 automatically get imaged? In the wrong hands such a decision instrument could increase the amount of chest imaging.

   I’m assuming that the gold standard for thoracic injury is chest CT. If so, shouldn’t the decision instrument be compared to the gold standard? In total, 44.6% of the 9905 enrolled patients had a chest CT. 43.1% only had 1 CXR, and 42.0% had both a CXR and chest CT. If less than half of the patients did not receive a chest CT, I’m assuming a certain percentage of TICIs may have been missed

   Do you agree with the definitions of clinically major, minor, and insignificant injuries? Is a hemo/pneumothorax only clinically significant if there’s a surgical intervention? What about pulmonary contusions?

   Several of the missed injuries also had tibia, pelvic and spinal fractures. At what point do you decide there’s a distracting injury to maintain such a high sensitivity?

   Does it matter that the study used convenience sampling between 7a.m. and 11p.m.?

Comments are closed.