Maitland, K et al.  Mortality after Fluid Bolus in African Children with Severe Infection. N Engl J Med 2011 June 30. 364(26): 2483-2495.

Compared to US and European practice, most hospitals in Sub-Sahara Africa do not routinely use fluid boluses for impaired perfusion in pediatric patients. This multi-center randomized controlled trial evaluated pediatric patients between 60 days and 12 years presenting severe febrile illness and impaired perfusion but no hypotension. Three groups of children were randomized 1:1:1 to either to receive boluses of 20 to 40 ml of 5% albumin solution (albumin-bolus group) per kilogram, 20 to 40 ml of 0.9% saline solution (saline-bolus group) per kilogram, or no bolus (control group) at the time of admission.  The primary outcome of 48-hour mortality occurred in 10.6% (111 of 1050 children) in the albumin-bolus, 10.5% (110 of 1047 children) in the saline-bolus and 7.3% (76 of 1044 children) in the control group. The trial was stopped early due to increased mortality rates in all bolus groups.  The primary limitation is that this is a different population with some different co-morbid conditions such as Malaria and HIV compared to US, but results did not differ despite these conditions in subgroup analysis. Author conclusions were fluid boluses significantly increased 48-hour mortality in critically ill children with impaired perfusion in non-ICU capable resource-limited settings in Africa. This study could have far reaching impacts for sepsis treatment in pediatrics as it directly refutes widely held dogma, but requires confirmation in US population of sick children with sepsis in urban and rural environment.

- Noel Hastings

Van Randen, A et al. The role of plain radiographs in patients with acute abdominal pain at the ED. The American Journal of Emergency Medicine. June 2011; 29: 582-589.

This prospective, multi-center study examined the added value of plain abdominal radiographs (XR) compared to clinical assessment alone in the evaluation of patients presenting with acute onset of abdominal pain.  From 3/2005 to 11/2006, a total of 1,101 patients were enrolled, age≥18yo, presenting with abdominal pain for >2 hours and <5 days, excluding pregnant women, patients with blunt or penetrating trauma, patients in hemorrhagic shock and patients discharged from the ED without any diagnostic imaging. All patients received a plain supine abdominal XR and upright chest XR. Patients were evaluated by ER (26%) or surgery (74%) residents who noted a primary diagnosis and confidence score both before and after acquisition of XRs. All patients then underwent imaging with either CT or ultrasound, and an independent expert panel subsequently assigned a final diagnosis after a chart review at 6 months. Changes in the overall accuracy of the initial diagnoses were evaluated for statistical significance. The pre-XR and final diagnoses corresponded in 49% of cases while the post-XR and final diagnoses corresponded in 50% of cases, a non-significant difference. A change from pre- to post- XR diagnosis occurred in 11% of cases, of which only 22% converted to the correct diagnosis. Among patients for whom the primary diagnosis remained unchanged, the level of confidence in the diagnosis was unchanged in 65% of cases. Bowel obstructions had a correct primary diagnosis in 57% of patients pre-XR and 74% of patients post-XR, which showed a p<0.01 for change in sensitivity. The study concluded that the added value of plain XRs in the diagnostic workup of abdominal pain is too limited to advocate for routine use. Limitations included the lack of decubitus abdominal views, and the lack of use of radiologist reads to guide diagnosis. Of note, plain abdominal XRs may still be useful for expeditious diagnosis of bowel perforations or foreign bodies, as well as to guide preparation for subsequent imaging such as oral contrast for CT in the evaluation of bowel obstructions. It would, in addition, be helpful to know what percent of patients presenting with clearly diagnostic plain XRs for free air or bowel obstruction actually have surgical intervention performed prior to or without obtaining CT imaging.

- Mike Niedermeier

Yannopoulos D, et al. Sodium nitroprusside enhanced cardiopulmonary resuscitation improves survival with good neurological function in a porcine model of prolonged cardiac arrest. Crit Care Med 2011 Jun;39(6):1269-74.

The use of epinephrine in prolonged cardiac arrest has not shown improved outcomes. The use of nitroprusside has been theorized to decrease arterial vascular resistance thus improving forward blood flow with chest compressions. This randomized, blinded, animal study in a preclinical animal laboratory used 24 female farm pigs (30 +/- 1 kg), and looked at sodium nitroprusside -“enhanced” CPR (SNPeCPR) compared to standard CPR (S-CPR) and enhanced CPR (eCRP - CPR + impedance threshold device + abdominal binding) in animals after cardiac arrest to assess effect on 24-hour survival rates. Twenty four Yorkshire female farm pigs (30 +/- 1 kg) were sedated, intubated, and monitored. Ventricular fibrillation was induced via a temporary pacing wire and ventilator was disconnected. After 8 minutes of untreated ventricular fibrillation, 24 pigs were randomized to 25 minutes of S-CPR (n=8), eCPR (n=8), or SNPeCPR (n=8). The S-CPR groups received 0.5 mg boluses of epinephrine every 5 minutes. The SNPeCPR group received 1 mg boluses of SNP every 5 minutes and did not receive any epinephrine. Defibrillation was delivered after 25 minutes of CPR. Animals that had return of spontaneous circulation (ROSC) were observed under general anesthesia until hemodynamically stable. At 24 hours, a certified veterinarian blinded to the intervention assessed the pig’s neurologic function on an overall performance category (OPC) scoring system (1=normal; 2=slightly disabled; 3=severely disabled but conscious;4=vegetative state; 5=dead). No animal treated with S-CPR with epinephrine had successful ROSC and therefore could not be assessed for 24-hr survival. 8 of 8 animals that received SNPeCPR had a ROSC, and 7 of 8 survived for both 1 and 24 hours. 5 of 8 animals with eCPR had an initial ROSC; 2 of 8 survived to 1 hr and 2 of 8 survived 24 hours (p = 0.04 for 1-hr and 24-hr survival compared with SNEPeCPR.)  Good outcomes at 24 hrs (OPC 1 or 2) were significantly better in the SNPeCPR group than the eCPR group (6 of 8 versus 1 of 8) (p=0.04).  These results suggest the feasibility of using SNP in treatment of cardiac arrest to enhance brain flow and increase 24 hr survival with favorable neurologic outcomes. Limitations of this study include the lack of investigation into the biochemical mechanism responsible for improved outcomes with SNP (presumed to be secondary to vasodilatory effects or theoretically to protective effects of nitric oxide (generated by metabolism of SNP). Also of note, the study did not look at the combination of S-CPR + SNP. Also of note is the study treated animals in ventricular fibrillation with CPR without defibrillation for 25 minutes.

- Rachel Rockford