一起來討論journal article: "Imaging strategies for detection of urgent conditions in patients with acute abdominal pain"

這篇BMJ的文章滿有意思的，一共收了從六個醫院急診的1021個病人，每個都做plain radiographs, ultrasonography, and CT，table 2裡看到clinical diagnosis alone的sensitivity跟其他個別的diagnostic test是幾乎一樣的(statistically and clinically)，不過要注意的是這裡是要detect "urgent" etiology。雖然他們有稍加敘述如何用panel的來定義，覺得還是不夠詳細和精確。

他們花了很多的工夫給每個acute abdomen需要imaging的病人做了這三樣檢查，可惜的是沒對examiner和interpreter做好的standardization。很多的ultrasound & CT是在沒有supervision的情形下做檢查和判讀的，結果就是significantly lower sensitivity... 很有趣的是他們也有想到不要只針對全部的abdominal pain來做分析，於是在strategy 10 & 11就有比較；更有趣的是這樣的結果和CT alone是相近的，於是乎American College of Radiology guideline似乎就不是很有力了…

另外是對於這樣hierarchical的結構，也許是可以考慮multilevel的分析方式，當然只有六個醫院沒法做太詳細的分析，可是我覺得至少要用empirical Bayes的方法來分析… 再來是一個小東西，figure3的ROC space小弟覺得有點名不正言不順，since這裡的diagnostic是binary的，沒有任何的cutoff可以用，不認為是否可以仍然以ROC命名…

先野人獻曝一下，希望大家不吝提出高見… 下週我們的journal club也會討論，到時再來跟大家分享結果。

BMJ的網站做得很不錯，還有一堆空間給reader來做rapid response...

ED grand round 11/13/09

Again, I join the Hopkins weekly grand round today. Along with the residents candidate interviewing to day, many attending physicians join the meeting today, which made the meeting more interesting. Just again share my note with everybody who might be interested.

Today the topic is Sedation

Pharmacology

Importance of ETCO2:

CO2 decrease before O2 (more sensitive and earlier)

Brady = hypoxia
Tachy = hypercapnia until proven otherwise
Hypoxia sequence (EtCO2 down, SpO2 down, then Bradycardia)
Cyanosis: happens when SpO2 <>

Advocate the usage of Propofol
Propofol 20+% hypotension, bolus only otherwise redistribute propofol
infusion syndrome lactic acidosis

(Hopkins have yet started the usage of Propofol in ED, but will be soon.)

One person does procedure one does sedation (will be policy)

This is not new to me. At my visit in U of Alberta, I realized that's a two man procedure. When can we really start to do EM @ TW sophisticated like this?

Fasting rule 8hr solid food 4 breast milk 2 clear liquid (policy)

www.Hopkinsinteractive.org (everyone needs to have certificate)

will share more information after I visit the web

Midazolam peak in 2-5min whoch might be the reason over
Mike never use that to sedate unless intubated already

(quite different from the practice @ CGMH back in TW)
Diazepam is quicker than mid (oral is opposite) he thinks safer drop blood pressure

Barbitals
Methohexital barbituate peds like but not great for RSI though it's
quick onset and offset
Pentobarbital longer
Thiopental longer and quick safe

Etomidate is better (widely used drug for RSI in US)

Choral hydrate oldest great for peds slow onset last hrs t1/2 10hrs

Etomidate has myoclonus effect, however some say transient. adrenal
insufficiency is owing to prolonged use. NEJM article (for randomized trial)
Propofol has musclelysis (do we have this term?)

Ketamine in peds iv im (the reason why ppl loves it)

Nitrous oxide for dentist (in his rotation)

Single vs mixed drug use debate eg ketofol

Hopkins RN or not creditential personnel can't give deep sedation agents

High risk pt want to involve anesthesia
Previous problem
Extreme of age
Sleep apnea
Obesity
Pregnancy
Multiple trauma
Substance abuse

ASA category
Evaluate
Discharge plan

Cuff20-50% should be greater than arm circum

Paradoxical motion
SpO2 -30 = PaO2

Head position picture shoulder should be higher (different airway structure)

Sequence of getting Epi or atrop depends on availability (brady get epi first then atropin in peds)

Add bicarbonate in local (1:9 in lidocaine)

Simulation room:
same situation like ACLS different from TW: give compliment before criticize

Meningitis:
Nuchal rigidity only has 30% sensitivity
Burski sign even worse
the head rotation test has 93% sen, of course, not specific
chemoprophylactis done throughly in there
steroid need to be given
worry about the Listeria for >50, immunocompromised, etc. give ampicillin
RBC found in non-traumatic tap, worry about HSV. tap before given acyclovir

Chairman chat again panel type
Synco-pe

again they use the interactive remote answer card
www.turningtechnologies.com

http://www.turningtechnologies.com/professionalaudienceresponse/audienceresponsesolutions/
Head or facial trauma hints the duration or severity of cause of syncope
Open discussion is interesting but need the culture to do that
TnI for risk stratification

Interesting research forum today

Research forum is the monthly dialogue we have here in GTPCI at Hopkins. Students are asked to either share their research dilemma or initiate discussion the might be helpful for their project in this forum. Today a surgeon shared the project that makes him able to jump up to wear his shoes every morning about the decision that physicians or surgeons make for the resection of liver metastasis from colorectal cancer.

It's interesting because this surgeon, although very passionate, is a sort of 'typical' surgeon. By saying that implies the stereotype of surgeon is confident but sometime arrogant. Today he just keeps giving the example that he operated a 35 year-old having 4 kids man, who has been told by another doc "to have a cruise". Obviously there's some gap between the society of surgeon and medical oncologist. How physicians make decision is going to be his main research question. But since he who's also in the panel of expert to make the guideline thinks the consensus now is 'arbitrary', seems the director's suggestion: go make the guideline more solid by doing some outcome analysis is very wise. However, this 'typical' surgeon seemed not able to pick up his idea but still trying to defend himself...

Very interesting research forum today...

JHU ED 週會筆記 101609

今天的週會滿有趣的，有些是炒冷飯和一些小的收獲的就不在本Blog贅述(像是resuscitative thoracotomy)，重點是這個新的chairman's panel時間。原本這時間是J大的Chair在唱獨角戲配合幾個intern演出，自從resident director換人後有些改變，所以今天這時間改成一個panel在台上討論，好幾位資深和資淺的主治醫師們在台上分享他們的Differential和處理的原則。今天的case比較偏向medical legal issue，國情不同也不拿出來長篇大論，主要是這樣型式的會議滿有意思的，也許大家可以考慮試試…

Cheating death

平常上下學(班)都是聽NPR，一方面關心美國生活一方面練英文。

今天fresh air訪問的Dr. Sanjay Gupta是個神經外科醫師，寫了一本新書叫"Cheating death"(要翻成欺騙死亡嗎？)。上電台講了幾個hypothermia治療的個案；第一個是個年輕女孩，在低溫下環境DOA(還沒搞清楚現在叫什麼名字)，反正就是medics也沒有看到任何vitals, 不過按照hypothermia的原則，他們並沒有在現場declare death(這是美國paramedic可以做的事)，也沒有在現場回溫就送到醫院。到院後也沒有太多Aggressive的治療，病人還是沒有任何vitals，遵照hypothermia的流程，緩慢的回溫(不好意思，因為這是public radio，他沒給太多細節)，沒有給任何iv fluid(因為會跑到extracellular)，後來病人不止活了過來，還考上醫學院，到救活她的醫院服務。不過這本書另外的重點是deep coma的病人，什麼時候才要放棄他們？當然，聽到熱血是因為這個hypothermia的治療很多research在進行，而這樣的field才是我想要做的…

有興趣的人可以從下面的連結去聽聽這段訪問…

http://www.npr.org/templates/story/story.php?storyId=113681104

Cadaver's Lab

知道J大急診有在做這個lab一陣子，終於今年有機會先往一睹，也磨磨久沒用的老刀…

差點被traffic害得去不成的時候，還是看到不守時的美國人在門口等人，所以好家在老狗我也才得以進去…

還是稍稍解釋一下這個lab到底是做啥的：與一般大体解剖不同的是，這是個給臨床醫師訓練的大体實驗室。其實外科的同學們可能就比較有經驗/概念，主要就是在大体上練procedure。不過對急診醫學界這還是滿新的概念，手上是沒有相関的data，不過這樣難得的經驗，有許多醫學系或是Physician assistant的學生intern們都擠著來就可以看得出其空前的程度了。

一具大体大約是有四到五個學生在學習，每具當然都有一位主治醫師指導。開始就從cvp的不同下手方式教起；不過強調的是IJ(internal jugular vein)現在他們都一定會使用超音波定位才下手了，這滿值得參考的。CVP在大体上學習並不是最適合的，因為不容易看到回血。最適合的幾個procedure大概是cricothyroidotomy, tracheostomy, tube thoracotomy, and thoracotymy。我們這具大概是第一個開胸的，隔壁組也跑來看。其實講來這個procedure真的很簡單也很快，因為目的就是要快啊，不然就沒意義了；重要的還是適應症的拿捏吧。

能在大体上練習這些procedure再到真人身上訓練是比較好的，雖然現在很多假人模具可以擬真到一定程度，有些的anatomy還是不能做到的；至於來源的問題，也有可能是能不能在台灣做的関鍵之一了。

JHU ED 週會筆記 M&M

今天睡了過頭，不知想啥昨天鬧鐘設定好但沒切到開關，所以只來得及聽M&M，不過也有些精彩的case...

會議一開始是院方的律師來解說依照 Joint Commission on the Accreditation of Healthcare Organizations (JCAHO)訂定的Sentinel event report process；依規定有許多是要通報Maryland州政府的事件，去年在推動的其中一個便是wrong site surgery。以前在本blog有提過的，美國醫界為了減少medical error推動讓大家可以把錯誤講出來而不是pointing fingers的文化，在今天的會議中充份顯示出成果。想想我們本島還有很長一段路要走，從北城醫院推護士出來，邱小妹事件推住院醫師出來，都有點不是真正處理事件而是找人頂罪的味道。

講完SE的報告後，就要提今天的重點了，六個'04年以來的重大RCA(Root Cause Analysis)報告；當然每一個RCA之後都有所改進。第一個case是一個walk-out MI，病人主訴10/10 chest pain，with radiation to right shoulder。病人在檢傷等了約一小時後離開，到其他醫院急診發現是AMI。這個case後來伴隨的改進就是這樣的病人在檢傷就自動開始要在10分鐘內做好ECG並交給主治醫師判讀；知道院內最近也開始推動這樣的作法。第二個case則是一個十二歲騎單車被撞的facial trauma，在急診時GCS滿分，血壓不穩定。急診花了一堆時間處理airway都沒有成功，後來就只好做cricothyroidotomy，病患死於做cric的失血過量。從這個case以後就regular舉辦airway workshop，並成立airway team24/7隨時on call處理difficulty airway。另外一個教訓是從此以後做cric的incision一律不准做horizontal的…

剩下四個case都是嚴重且最常見的communication problem。且挑一個最近的case來講：這是一個neck squamous cell carcinoma的病人，來診主訴疼痛。檢傷紀錄就這麼檢單，不過看到醫師的note裡有一個十分可疑的地方：respiratory rate 24。病人於下午近六時就診，在八時至九時打了四劑的dilaudid…這時候lab data也回來了，BUN130, Cre3.6, Lactate 7.9；針對這樣的data在chart上看不到任何的處置。差不多快十一點了，這個灰姑娘時間跟本院是一樣的，sign off的時候到了。護理記錄在這之間也沒月任何特別的呈現，就還是差不多的vitals，不過醫師認為有可能是sepsis，開打了vancomycin等antibiotics。不過交完班過了一個多小時之後，護理記錄上出現了tachycardia, dry mucous, patient unconscious，這時候才看到bolus fluid的紀錄，病人最後撐不過大夜班就走了…

紀錄上寫得那麼少，真的醫師們都沒做夠嗎？一方面是紀錄不足的問題，不過上法院就只看你的文字了…另外就是，醫師辯稱原本要讓病人住ICU的，無奈樓上住院醫師一句話說：不需要，就只好去找monitor bed了。這一點之前就在週會討論過了，JHU和長庚系統一樣，是少見24小時都有board certificated急專醫師註診的醫院，大伙氣不過的是這樣board certificated的專科醫師的order竟然給junior resident就可以否決了？Chair馬上就補充這一點其實從1995年就有背書了，因為這樣的急專主治醫師24小時註診，JHU要求這樣的對話要發生在主治醫師之間。Dr. Kirsch馬上也補充他個人近兩個月的經驗，也是要讓病人住ICU的septic shock病人，樓上住院醫師就自己決定住monitor bed即可，甚至還辯稱已和VS討論；於是Dr. Kirsch真的打給主治醫師，當然就完全不知道這個case且馬上讓病人住上ICU了…

當然我也在facebook上看過一個學弟抱怨急診半夜收了一個住一兩天就出院的病人，至於大夜要不要上病人是另一個問題。不過這個case發生之後，大家熱烈的討論，最多的還是關於急診紀錄，以及交班容易出現的問題。許多attending分享自己處理的方式，大部份就是交班之後要重新審視病歷及病人；關於這點JHU EM裡有個team正在做整個配套的設計，等有新消息再上來分享…