sinaiem.org

Thanks to Mieka, Liz, Dr. Beattie, Dr. Weingart, and our Grand Rounds speaker, Dr. Sharma, for presenting early this Wednesday morning. Seems like a lot of us were up late for a Tuesday night, but those that made the early morning trek to the Hurst were well rewarded.

Mieka’s case report and talk on Sgarbossa’s criteria for AMI in LBBB provided a great overview of this controversial topic. The criteria are below, and I’m reprinting the findings from an EM meta-analysis in last month’s Annals (Tabas et al, Volume 52, Issue 4, October 2008, Pages 329-336,  PMID: 18342992):

The original Sgarbossa paper (NEJM 1996) awarded points for the following findings on EKG, and determined a positive likelihood ratio of 7.8 for patients with 3+ points, and a negative LR of 0.2 for those with less than three points:

5 points for:  ST-segment elevation >1 mm in lead with concordant QRS complex
3 points for:  ST-segment depression >1 mm in leads V1, V2, or V3
2 points for:  ST-segment elevation >5 mm in lead with discordant QRS complex

Different groups have found different sensitivities over the years, mostly low — so using Sgarbossa to rule out the need for reperfusion therapy was always a dicey proposition. The conclusion of the Tabas meta-analysis was:

Our results demonstrate that a Sgarbossa ECG algorithm score greater than or equal to 3, defined by at least 1 mm of ST elevation that is concordant with the direction of the major QRS or 1 mm of ST depression in V1 to V3, has a high specificity (98%) and a useful positive likelihood ratio (7.9). This positive likelihood ratio is the same as or higher than those for 7 of the 10 ECG criteria currently recommended to diagnose acute myocardial infarction in patients with normal interventricular conduction, including ST depression, T-wave changes, or presence of any q waves.³⁴ The predictive value of the Sgarbossa ECG algorithm may also be significantly higher than the value of new left bundle branch block, which has been reported to have a positive likelihood ratio of only 1.4.20 M.C. Kontos, R.H. McQueen and R.L. Jesse et al., Can myocardial infarction be rapidly identified in emergency department patients who have left bundle branch block?, Ann Emerg Med 37 (2001), pp. 431–438. Abstract | PDF (106 K) | View Record in Scopus | Cited By in Scopus (24)²⁰ These data indicate that clinicians in appropriate clinical settings can confidently treat for acute myocardial infarction when a score of 3 is reached. Although the current American College of Cardiology?American Heart Association guidelines for treatment of patients with ST-elevation myocardial infarction discuss the Sgarbossa ECG algorithm score, they do not include recommendations for its use in treatment decisions of patients with left bundle branch block.³⁵ In light of the powerful positive likelihood ratio in our current report, consideration should be given to including recommendations for prompt revascularization when a score greater than or equal to 3 is found.

A Sgarbossa ECG score of 2, representing discordant ST deviation greater than or equal to 5 mm, is inadequate to diagnose acute myocardial infarction according to the results of this investigation, along with that of Sgarbossa et al.² In addition, given the poor sensitivity and poor negative predictive values of the algorithm, it is clear that their absence (a score of 0) does not rule out acute myocardial infarction.

Good to keep in mind, as LBBB patients make up less than 3% of our ED patients but account for 7% of AMI, and tend to do worse — possibly because we tend to miss them.

Dr. Rob Sharma’s talk on ED drug interactions was an excellent and timely review, plus he managed to highlight a new website sure to make its way into your bookmarks: qtdrugs.org (I am rating this site somewhere between extravasation.org.uk and aorticdissection.com). Some pearls to highlight:

• – The effect of Cipro on Coumadin (warfarin) is deadly — INR rises on day 4 or 5, up to a median level of 10. Levaquin doesn’t cause these problems. Use it instead.
• – Bactrim, macrolides and cephalosporins are other antibiotics that potentiate the effect of warfarin on INR. Rifampin antagonizes warfarin.
• – NSAIDs on a person taking a sulfonylurea like glyburide tend to promote hypoglycemia. (We prescribe NSAIDs more than anything else. If we ranked it, deaths from NSAID complications would fall somewhere between murder and AIDS).
• – Watch out for serotonin syndromes when giving Reglan (metoclopramide) to your patients on SSRI’s. Zofran and the ‘trons aren’t safe in this situation, either, though Compazine (prochlorperazine) is.
• – We saw a huge drop-off in ED droperidol use after the FDA’s black box warning. Now that Haldol carries the same warning, a new generation of residents, with nothing to lose, is bravely discovering the joy of droperidol. Just be sure to get an ECG after you’ve knocked out your patient (ideally, they should be on a monitor). More on this FDA chicanery here and here. Biros and Miner and company also published a great trial of droperidol vs. midazolam for undifferentiated agitation in 2005, unfortunately right after the ACEP guidelines came out.

Finally, with regards to the case of the missed rib fractures and PTX, I wanted to highlight a hypothetical but occasionally-real scenario: what if the PTX wasn’t missed, but in fact occured well after the injury? Delayed pneumothorax in the setting of blunt chest trauma and rib fractures is actually well-known phenomenon. My quick lit search found a good Greek paper on this topic, looking prospectively at over 700 patients with blunt trauma discharged and later found to have delayed PTX, HTX, or occult PTX (Misthos et al, Eur J Cardiothorac Surg 2004;25:859-864):

DPX was detected in 14 patients (2%) (Table 2). The most frequent associated injury with DPX was found to be one or two rib fractures (50%) that might have a causative relationship with DPX… Analysis of associated injuries and mechanism of injury in each entity did not reveal any reliable prognostic factor, but a strong correlation between rib fractures and DHX. Among all motor vehicle and fall accidents only 20 (3.2%) led to OPX and among those with chest wall muscle contusion only 22 (3.6%) developed OPX (Tables 1 and 2). DPX was described in 10 patients after fall or sports’ accidents (2.2%). It is important to mention that in 10 cases out of 42 OPX and DPX (23.8%) no other injury was traced except ipsilateral subtle chest discomfort at the end of deep inspiration. However, in 18.9% of all patients classified with no associated chest injury, OPX or DPX was detected (Tables 1 and 2). DHX was more frequently observed after motor vehicle collisions or pedestrian accidents (80.9%), while only 5.2% of such accidents developed DHX (Tables 1 and 3). At least one rib fracture was present when DHX was developed. Fifty-two patients out of 406 with rib fractures developed DHX (12.8%).

The authors suggest that blunt chest injury, even when not initially serious, might hide a lot of perils for patient’s life and surgeon’s reputation. Therefore, a routine follow-up for all patients is recommended irrespectively of the severity of initial findings.

No prognostic factor was detected in order to classify patients in different follow-up algorithms. All patients who do not fulfill the criteria for intrahospital management after minor blunt chest trauma should be closely followed for at least 2 weeks after the accident. This is supported by the facts that OPX, DPX and DHX should be anticipated, cannot be predicted and have excellent prognosis if detected in time.

Also, if a certain medical director ever pimps you about a supine CXR abnormality, be way of Deep Sulcus sign. Here’s an example.

Posted on Thursday, November 6th, 2008 at 5:42 pm by Nick. Filed under Trauma, Post-Conference Letter, Sedation, Arrhythmias, Toxicology, ACS, Blog.
You may post a comment.