d'Almeida KW, et al; for the Emergency Department HIV-Screening Group. Modest Public Health Impact of Non-targeted Human Immunodeficiency Virus Screening in 29 Emergency Departments. Arch Intern Med. 2011
This multicenter, cross-sectional study looked at the impact of non-targeted HIV screening in the ED setting. It was conducted in metropolitan EDs in Paris, France. Of 31 adult EDs in the French acute syndrome surveillance network (OSCOR), 29 agreed to participate in the study and were randomly assigned 6 week blocks to enroll eligible patients. Eligible patients were all patients aged 18-64 presenting to the ED and able to give informed consent (opt-in model). Exclusion criteria were self-reported HIV infection; inability to provide consent, because of neuropsychiatric disorder, substance abuse, language barrier, or being under arrest; unstable medical illness; and being seen in the ED for prophylaxis after sexual exposure to HIV. Of 138,691 visits, 78,411 patients were eligible, and 20,962 (27.0%) were offered screening at triage via a fingerstick, whole blood HIV-rapid test (HIV-RT), and 12,754 (16.3%) accepted screening. From this, 18 patients (0.14%, 95% CI 0.08-0.22%) were given new HIV diagnoses.  All patients were told their HIV-rt results (whether positive or negative) in the ED. Patients with positive HIV-rt results had confirmatory serologies drawn in the ED and were scheduled for 72h follow-up with in Infectious Disease (ID) clinic. Of the 18 patients diagnosed with HIV, 12 had previous HIV testing (median time since last HIV test 1yr) and 8 had late stage disease (either symptomatic HIV or CD4 counts <350/µL). However, all but one of these patients (n=17) belonged to a high-risk group (born in Sub-Saharan Africa [SSA] or men who have sex with men). Of the 18 patients with newly diagnosed HIV, 12 followed up in ID clinic, but 6 (33.3%) did not return for their first ID appointment (4 women, 2 men, all born in Africa).
There are several limitations of this study and its generalizability to other settings. While 20,962 patients were enrolled, it is unclear why nearly 75% of eligible patients were not offered HIV screening. Subgroup analysis of those not offered the HIV-rt screen (who they were and why was it not offered) would strengthen the study. Also, the study was done in a metropolitan center with a much higher rates of new HIV diagnosis than the rest of France (half of all new HIV diagnoses in France are in Paris according to the authors), arguing that there is either a higher per capita rate of undiagnosed HIV elsewhere in France, or that the marginal benefit from non-targeted screening would be even lower outside of Paris, arguably anywhere with lower HIV rates than Paris. Also, because 96% of the French population is covered by government-run health care (per Wikipedia), it is arguably easier to guarantee access to follow-up in cases of new HIV diagnosis than in the US. and elsewhere with a different healthcare system. Finally, the relative rarity of HIV in this population makes subgroup analysis of those who were diagnosed with HIV hard to interpret (though not specifically given, one would assume the confidence intervals for risk factors, etc are wide enough to make these factors not statistically significant). However, the results seem to argue that non-targeted screening finds very few individuals who would not have already been found with targeted screening (i.e. screening only in high risk groups).
In an era of recommendations for non-targeted screening for HIV from the CDC, as well as multiple other organizations worldwide, this study is the largest of its kind to date and raises several interesting questions. First, although the number of new diagnoses was small, all but one of these patients was considered high risk, raising the question, ‘Does the marginal increase in detection (1 additional case) warrant non-targeted screening over targeted screening?' From the data provided there were from 630-1166 individuals in the high-risk groups cited in the paper (if all MSM were from SSA and if none of the MSM were from SSA, respectively), meaning that between 11,588-12,124 were screened without these risk factors to find 1 additional case of HIV. Second, of 18 newly diagnosed HIV patients, 33% did not follow up, begging ‘Is the ED the place to do this screening?' While non-targeted screening done by PMDs may give increased opportunity for follow-up, the counter argument may also stand that many of our patients do not have a PMD, at least regularly (due to insurance status, immigration status, financial limitations, etc.) and if we are not screening them, nobody is.
Is this paper practice changing? I think not, but it is provocative and likely this issue of compulsory HIV screening will continue to be a subject of controversy, both in the public and in medicine, in the ED and beyond.

-Bryce Pulliam

Mollberg, NM, Wise SR, Berman K, Chowdhry S, Holevar M, Sullivan R, Vafa Al.  The consequences of noncompliance with guidelines for withholding or terminating resuscitation in traumatic cardiac arrest patients.  J Trauma 2011;71:997-1002.
This retrospective chart review of a level 1 trauma center's database evaluated the effectiveness of resuscitation of patients with traumatic cardiopulmonary arrest (TCPA) who were not appropriately managed using guidelines regarding termination or withholding of prehospital resuscitation.  The database included all trauma patients from 2003 through 2010, managed by the Chicago Fire Department, and who had a principal or admission diagnosis of cardiac arrest.  Only patients with a prehospital TCPA, who received CPR by BLS or ALS units, were included.  Further inclusion criteria were based on guidelines for withholding resuscitation, created by the National Association of Emergency Medical Services Physicians and the American College of Surgeons Committee on Trauma (NAEMSP/ACS-COT).  These were: 18 years or older; victim of blunt trauma found by EMS to be apneic, pulseless, and without organized cardiac activity by ECG; victims of penetrating trauma found to meet these criteria, with no other signs of life; patients not meeting these criteria, but who had greater than 15 minutes unsuccessful CPR (no ROSC) and/or transport time greater than 15 minutes after identification of TCPA.  Exclusion criteria were also based on NAEMSP/ACS-COT guidelines: younger than age 18; arrest secondary to drowning, electrocution, strangulation, hypothermia, or fire; those who may have had an atraumatic cause of arrest; patients transported by agency other than Chicago Fire Department; incomplete prehospital records; pronounced dead at the scene or en route; or with EMS witnessed TCPA but less than 15 minutes total prehospital time.
The final cohort consisted of 294 patients.  35% had signs of life on arrival to the ED, and 13% had ROSC upon presentation or during their time in the ED.  55% received ACLS medications in the ED.  5% were taken to the OR.  Of all of these patients, 2.7% (8) survived to the ICU.  4 were declared brain dead, 1 had withdrawal of care, and 2 died within 24 hours.  One patient (0.3%) survived to discharge to a long-term care facility with a GCS of 6.  There were no organ donors.  The total cost to the hospital for these 294 patients was almost $4 million, and the 20 patients who survived the ED accounted for 35% of this total.
This study reveals the effectiveness of following prehospital care guidelines for this patient population, and the extreme unlikelihood of positive outcome of these cases.  This is the largest study of its kind; multiple other smaller studies have had conflicting results.  The authors note that while another study of 138 patients, with 14 survivors (10%), states that 13 of these would not have been resuscitated if the NAEMSP/ACS-COT guidelines were followed, they utilized the guidelines incorrectly.  They required that ALL of the criteria be met in order to proceed with resuscitation, instead of just one.  If the guidelines were correctly followed, only 3 would not have been resuscitated; these patients had final GCS of 5, 9, and 10.  Given the apparent extreme unlikelihood of return to any semblance of functional neurologic status in this patient population, and the high immediate and long-term costs of care for neurologically devastated patients, it seems appropriate to use such a rule when applicable in the prehospital setting (and on arrival to the ED) to know when to proceed with medically appropriate care.  It would be helpful to know if the EMS crews in this area use these criteria, or others, to determine when resuscitation is appropriate; and if not, it may be appropriate to implement such guidelines.  The study is directly applicable to our facility.

- Daniel Naylor

 Klein, E., et al. A Randomized Clinical Trial Comparing Oral, Aerosolized Intranasal, and Aerosolized Buccal Midazolam. Annals of Emergency Medicine; 2011; 58:323-329
Emergency physicians sometimes administer oral midazolam for minor pediatric procedures not warranting parenteral sedation, but it is unclear if midazolam can be more effectively administered by buccal or intranasal aerosol routes.
This study is an ED based, single-center, randomized controlled trial occurring in an urban Children's Hospital ED comparing three routes: oral, aerosolized buccal, and aerosolized intranasal.  During a 3 year period, children aged 6 months to under 7 years requiring laceration repair whom were NPO for at least 2 hrs, had English-speaking parents, and for whom parents and the ED physician agreed that non-parenteral sedation was needed were approached for enrollment in the study.  Exclusion criteria included: children with oral/nasal wounds (which could impede drug delivery), closed head injury with LOC, abnormal neurologic exam, significant development delay or baseline neurologic deficits, severe trauma with suspected internal injuries, acute or chronic respiratory, renal, cardiac, or hepatic abnormalities, known allergy or previous adverse reaction to benzodiazepines, use of an erythromycin-containing antibiotic (which could affect benzodiazepine metabolism), or previous enrollment.
Primary outcome was the first CHEOPS score after the laceration repair procedure began.  The CHEOPS (Children's Hospital of Eastern Ontario Pain Scale) is a 4- to 13- point scale that measures cry, facial expression, verbal expression, torso movement, leg movement, and reaching toward the wound (with lower scores indicating less distress). The CHEOPS score at the beginning of the procedure demonstrated modest, but statistically significantly, less distress in the buccal group compared with the oral group, but no difference between the intranasal and oral groups.  Despite the primary outcome, secondary outcomes favored the intranasal route with demonstrated faster sedation onset, greater proportion of patients achieving adequate sedation, a greater proportion of patients with optimal activity scores, and a higher rate of parents who would choose the same regimen again.  Unfortunately, the intranasal route also demonstrated the most immediate irritation, and smallest proportion of patients who accepted the medication easily.
Limitations include subjective ratings of "adequate sedation", and delay from time of adequate sedation to procedure start time in all groups.
Overall, I feel the impact that this study, while applicable to our patient population, has minimal impact as the authors did not find one route to be definitively superior to the others.  It does demonstrate that aerosolized buccal or nasal administration is at least as effective as the oral route.  It is important for the ED physician to remember that there are a variety of options when individualizing a patient's sedation. 

- Christopher Ang