Sise MJ, Kahl JE, Calvo RY, et al. Back to the future: Reducing reliance on torso computed tomography in the initial evaluation of blunt trauma. J Trauma Acute Care Surg. 2013 Jan;74(1):92-9. A trauma surgery group at an academic level one trauma center performed a pre/post intervention comparison with the primary outcome being the rate of chest/abdomen/pelvis CT for trauma activations. During the initial phase they retrospectively gathered CT utilization rates for the different trauma surgeons as well as their patient and injury data. Reportedly no surgeons had diagnoses errors or missed injuries, so they took the surgeon with the lowest CT utilization rate and essentially the whole surgery team agreed to adopt their practice pattern “a de-emphasis on the mechanism of injury combined with an increased reliance on the findings of the physical examination and FAST...” After this change in practice by the group they then prospectively collected similar data on CT utilization, patient demographics, injury pattern, and mechanism.

They reported about 900 trauma patients in each cohort, and they generally had similar demographics, injury pattern, and mortality. All physicians dropped their CT utilization rate - it decreased by 38.5% overall and there was a more than two fold increase in FAST and plain chest and pelvis films. There were significant quantities of data presented correlating demographics and injury pattern to likelihood of having received imaging in the two cohorts but this was generally of very unclear importance or of poor statistical significance - or both. Again, no “errors or missed injuries” were reported - seemingly justifying this somewhat haphazard reduction of CT utilization in the eyes of the publishers.

I applaud the quest for individualized patient selection to imaging modalities that use ionizing radiation. With that said I am quite surprised that a major academic group would publish such a report, as there seems to be numerous shortcomings which include:

  • unclear what practice pattern was adopted or how they measured adherence
  • the person who the pattern was molded after had a significant drop of imaging utilization - either because they knew they were being studied or because the cohorts differ
  • poor process for identifying missed injuries
  • numerous exam findings well documented in EM literature to modify the pretest probability of important injuries are not accounted for

Overall a well-intended study that has significant methodological deficiencies and does not address the current literature on imaging in trauma

- Cotton 

Downar J, Kritek PA.  Family presence during cardiac resuscitation. N Engl J Med. 2013 Mar 14;368(11):1060-2. Family presence during CPR is generally accepted to be beneficial to family members, but few studies have actually looked at this issue.  This is a French multi-center, randomized, controlled trial of first degree relatives of patients who had cardiopulmonary arrest at home and underwent CPR.  Fifteen pre-hospital EMS units including driver, nurse, and EM physicians in France participated.  Half of the units were randomized to ask family members whether they wished to be present during the resuscitation (intervention group), and the other half followed standard procedure, which did not include offering for family members to watch.  79% of family members in the intervention group actually witnessed CPR, compared to 43% of those in the control group.  Ninety days after resuscitation one family member for each patient was surveyed by phone by a psychologist who was unaware of the study group assignment using the Impact of Event Scale and the Hospital Anxiety and Depression Scale.  The primary endpoint was PTSD based on based on scoring with these scales.  Secondary endpoints were effect of family presence of medical efforts, well-being of the health care team, and occurrence of medical legal claims, obtained by survey of medical team after resuscitation and report of medical claims by the hospitals.

570 family members were enrolled, 60% of whom were present for resuscitation.  Group characteristics were similar.  17% were lost to follow up.  Using intention to treat, those who did not observe CPR had an OR of 1.7 of endorsing PTSD related symptoms at day 90, with similar results for observed-case calculation (OR 1.6).  <1% of family members were aggressive or in conflict with the medical team.  12% of those who did not witness CPR wished they had, whereas 3% of those who did wished they had not.  There was no significant difference in stress levels of medical professionals with or without family watching.  No medico-legal claims were filed.  Effectiveness and duration of CPR and selection of drugs were not affected.  In conclusion, this randomized, controlled trial demonstrates that family members who were present for CPR were less likely to experience PTSD symptoms, and family presence confers no effect on medical efforts, well-being of the health care team, or occurrence of medical legal claims.

- Kellenbeck