Think before pulling the trigger: HTN Urgency/Emergency

Hi,

For this week we have a quick one, on Hypertension (Symptomatic/Asymptomatic) Management in the ED looking at HTN Emergency/ HTN urgency.

Asymptomatic HTN aka "HTN Urgency"

We get really worried about the numbers when we come across such patients with markedly elevated pressures who are asymptomatic! Well, You will just do fine if you perform a good history and physical on them, give them some time and recheck BP after sometime. Don't push them to the ICU just because of those numbers.

Lets look at ACEP 2013 policy on Asymptomatic HTN, They looked at 2 specific questions:

1. In ED patients with asymptomatic elevated blood pressure, does screening for target organ injury reduce rates of adverse outcomes?

(1) In ED patients with asymptomatic markedly elevated blood pressure, routine screening for acute target organ injury (e.g., creatinine, urinalysis, ECG) is not required.

(2) In select patient populations (eg, poor follow-up), screening for an elevated serum creatinine level may identify kidney injury that affects disposition.

2. In patients with asymptomatic markedly elevated blood pressure, does ED medical intervention reduce rates of adverse outcomes?

1) In patients with asymptomatic markedly elevated blood pressure, routine ED medical intervention is not required.
(2) In select patient populations (eg, poor follow-up), emergency physicians may treat markedly elevated blood pressure in the ED and/or initiate therapy for long-term control. [Consensus recommendation]
(3) Patients with asymptomatic markedly elevated blood pressure should be referred for outpatient follow-up. [Consensus recommendation] 

Sometimes doing nothing is adequate as about 1/3 of patients had lowering of BP in the ED with no medical intervention. So, when dealing with a asymptomatic hypertensive next time, don't get worried with the numbers, treat only if the patient is symptomatic. And it also equally important to educate them about the harms of rapidly lowering the BP. 

Symptomatic HTN

Hypertensive emergency is any form of acute end-organ dysfunction.
End organ dysfunction includes ARF, Eclampsia, Heart Failure, Dissection, ACS, CVA, Encephalopathy. (Don't look at specific numbers to diagnose HTN emergency!)

If you have someone with symptomatic HTN (SOB, Chest Pain, AMS, headache, focal deficits), first and foremost - try to find out the cause of elevated BP:

1. Ask them about medication compliance
2. Substance abuse (cocaine, amphetamines)
3. Symptoms related to Pheo, Thyroid, Renal Artery Stenosis, Coarctation of Aorta.

This subset of patients needs aggressive reduction of BP with IV medications (reduce MAP by 20-25% over 1 hour). Majority of them are volume depleted, so give them some IV fluids too.

Now the question is, which should be our drug of choice ?
It depends a lot on your history and working diagnosis because anti HTN meds work best if they are chosen with respect to the cause of elevated blood pressure. I do see people grabbing NTG for undifferentiated hypertension (regardless of the cause/symptoms), using venodilating wimpy doses, with which I don't really agree. 

DOC is based on your diagnosis, for instance, use:

• NTG - Acute Pulmonary Edema
• Labetalol/Nicardipine - CVA
• Beta blockers - Dissection
• Eclampsia - Mg/ labetalol - Eclampsia

Key Points:

1. Don't get worried with numbers, treat only if the patient is symptomatic. Asymptomatic HTN goes home with good follow up/discharge instructions.
2. High BP + end organ damage is HTN emergency, again don't look at specific numbers to call it HTN emergency.
3. Reduce MAP by 20-25% in the first hour with titrable IV meds (Exception dissection, eclampsia where you need to get down as quickly as possible)
4. Most of them are fluid depleted and need IV fluids, which also prevents the sudden drop in pressures after commencing IV therapy.
5. NTG drip is not the answer to everything! 

For further reading:

1.  Wolf SJ, Lo B, Shih RD, et al. Clinical policy: critical issues in the evaluation and management of adult patients in the emergency department with asymptomatic elevated blood pressure. Ann Emerg Med. 2013; 62: 59-68.
2. Marik PE, Rivera R. Hypertensive emergencies: an update. Current Opinion in Critical Care 2011;17(6):569-80. 
3. http://www.acep.org/workarea/DownloadAsset.aspx?id=90154

Cognitive pills for Cognitive ills: Errors in Emergency Medicine

Cognitive errors underlie most diagnostic errors that are made in the ED. And with time I have realised that our speciality is a vulnerable one, because we often commit some errors (delayed diagnosis, missed diagnosis, unnecessary imaging) which look like no-brainers to people upstairs in the ICUs and wards. Then why do we commit such errors?

May be because as everyone else, we often have a tendency to pursue more readily attainable goals.

There is a story about a jogger who came across a man on his knees under a streetlight one evening. He explained that he had dropped his wedding ring. The jogger offered to help him search, and he accepted. With no luck after a half hour, the jogger asked the man if he was sure he had dropped the ring at the place where they were searching. The man replied that he actually dropped it several yards away in the shadows. ‘‘Then why are we looking here?’’ asked the jogger. ‘‘Because the light is better,’’ came the reply. 

This is a topic in medicine which is rarely talked about, despite being a really important one for patient safety. These errors are universal but since we are often the first responders as Emergency Physicians, the brunt falls on us and we face the limelight. So, we need to find out a way to minimise these errors by developing a conceptual framework and strategies in this critical aspect of patient safety.

COGNITIVE ERRORS

Diagnostic errors arising through cognitive errors are those that are associated with failures in perception, failed heuristics, and biases are referred to as cognitive dispositions to respond (CDRs). There are a number of strategies for reducing them (‘‘cognitive debiasing’’)  like METACOGNITION, a reflective approach to problem solving that involves stepping back from the immediate problem to examine and reflect on the thinking process.

Some unique operating characteristics of ED predisposing to medical error:

• High Diagnostic Uncertainity
• High Decision Density
• High Cognitive Load
• High level of activity
• Inexperience
• Interruptions and Distractions
• Shift Work
• Shift Changes

There is huge list of errors which we can come across as clinicians and not surprisingly, all of them are evident in Emergency Medicine, a discipline that has been described as a ‘‘natural laboratory of error.’’ Lets familiarise ourselves with some of them:  

Various biases leading to errors:

1. Anchoring Bias: Anchoring bias causes physicians to stay with their initial impression of a case and fail to adjust to new information that would make the initial impression less likely. This often leads to prematurely ending their search or premature closure.
2. Gender Bias: the tendency to believe that gender is a determining factor in the probability of diagnosis of a particular disease when no such pathophysiological basis exists. Generally, it results in an overdiagnosis of the favored gender and underdiagnosis of the neglected gender.
3. Availability: Recent experience with a disease may inflate the likelihood of its being diagnosed. Conversely, if a disease has not been seen for a long time (is less available), it may be under diagnosed.
4. Premature closure: a powerful bias accounting for a high proportion of missed diagnoses. It is the tendency to apply premature closure to the decision- making process, accepting a diagnosis before it has been fully verified (When the diagnosis is made, the thinking stops)
5. Search satisfying: reflects the universal tendency to call off a search once something is found. Comorbidities, second foreign bodies, other fractures, and coingestants in poisoning may all be missed. 
6. Ascertainment bias: occurs when a physician’s thinking is shaped by prior expectation; stereotyping and gender bias are both good examples.
7. Fundamental attribution error: the tendency to be judgmental and blame patients for their illnesses rather than examine the circumstances that might have been responsible. In particular, psychiatric patients, minorities, and other marginalized groups. 

Few Cognitive de-biasing Strategies to Reduce Diagnostic Error Develop insight/ awareness: Provide detailed descriptions and thorough characterizations of known cognitive biases, together with multiple clinical examples illustrating their adverse effects on decision-making and diagnosis formulation. Consider alternatives: Establish forced consideration of alternative possibilities. Encourage routinely asking the question: What else might this be? eg: any pt who presents with flank pain/hematuria, force yourself to consider aortic dissection. Metacognition: Train for a reflective approach to problem solving: Metacognition is the process of actively stepping back from the pushes and pulls of the immediate situation (de-anchoring), reminding oneself of the limitations and failings of memory, seeing the clinical problem in a wider perspective than that dictated by the obvious presentation, perhaps reminding oneself of specific lapses or failures in the past, and finally activating known cardinal rules or caveats.   Decrease reliance on memory: Improve the accuracy of judgments through cognitive aids: mnemonics, clinical practice guidelines, algorithms, hand-held computers. Simulation: Develop mental rehearsal, ‘‘cognitive walkthrough’’ strategies for specific clinical scenarios to allow cognitive biases to be made and their consequences to be observed. Construct clinical training videos contrasting incorrect and correct approach. Make task easier: Provide more information about the specific problem to reduce task difficulty and ambiguity. Make available rapid access to concise, clear, well-organized information.Write the ddx in chart upon initial evaluation and re-visit the ddx when initial tests are back and when deciding disposition. Formalizing Handover: 'S BAR' Mneumonic for Handover - Situation, Background, Assessment, Recommendation Minimize time pressures : Provide adequate time for quality decision- making. Have the the attending doc and and the handover doc seeing patient/imaging together, its always better to review the H&P for handed over cases.  Feedback: Provide as rapid and reliable feedback as possible to decision makers so that errors are immediately appreciated, understood, and corrected, resulting in better calibration of decision makers. Understanding "how we think":  Type 1: The Intuitive/Reflexive Approach involves automatic decision making based on pattern recognition. It's fast, requires little effort and usually brings you the correct diagnosis, but it's very prone to error. Type 2: The Analytical/Problem-Solving Approach is more critical and logical. This is when you step back and think more carefully about the patient's presentation. It involves estimating pretest probabilities, continuous self-questioning, and considering alternative diagnoses. While it takes more effort, more time and is more resource intensive, it's reliability is much better than the intuitive approach, and is more likely to give you the correct diagnosis.  High Risk Situations where errors are likely: Night shifts, during handover, with patients at the extremes of age, the 'difficult patient' and the "difficult relatives"

Some Classic Errors:

1. Failure to consider a closed-head injury in an intoxicated patient
2. Incomplete consideration of AMI mimics before initiating thrombolysis 
3. Inadequate assessment of immunocompromise status in patients with animal bite wounds
4. Failure to fully assess the medical status of psychiatric patients before transferring to a psychiatric facility
5. Failure to consider tetanus immune status in patients with open wounds. 

Take Home:

1. Learn, practise and teach Metacognition.
2. Develop your own strategies to reduce errors (discussion, checklists, Incorporating simulation)
3. Use a Problem Solving approach instead of a Reflexive Approach.

For Further Reading:

1. http://www.cjem-online.ca/sites/cjem-online.ca/files/pg271(2).pdf
2. http://home.comcast.net/~jasoncillo/Cognitive.pdf
3. http://www.ncbi.nlm.nih.gov/pubmed/12915363 

The other "ACS": Abdominal Compartment Syndrome!

Hi, This week I have a podcast on ACS --> Abdominal Compartment Syndrome!

As Emergency Physicians we should be familiar with issues requiring the ED/ICU care. And now, Intra-abdominal hypertension (IAH) leading to abdominal compartment syndrome (ACS) are increasingly recognised in the ICU, often associated with abdominal trauma and in patients requiring fluid resuscitation, such as those with septic shock, major abdominal surgery, burns, bowel obstruction, ascites, etc. ACS is the progression to end-organ dysfunction if IAH is not recognised and treated. 

Lets listen to the podcast to find out some more details, Check out the show notes too!

ACS

Influenza - Who needs Oseltamivir?

Everyday I come across several patients who walk into the ED asking for testing Influenza virus (flu) and getting a vaccine. So I thought we should review this one, recent guidelines from Ministry of Health and Family Welfare on Influenza discussing who needs to be tested/treated/admitted and vaccinated in addition to some basics.

History:

• Swine flu is a respiratory disease caused by the influenza viruses that infect the respiratory tract of pigs, the virus can be transmitted to humans.
• Swine flu viruses may mutate (change) so that they are easily transmissible among humans.
• Swine influenza is known to be caused by influenza A subtypes H1N1, H1N2, H2N3,  H3N1, and H3N2.
• Investigators decided the 2009 so-called “swine flu” strain, first seen in Mexico, should be termed novel H1N1 flu since it was mainly found infecting people and exhibits two main surface antigens, H1 (hemagglutinin type 1) and N1 (neuraminidase type1). The present flu virus in India is A(H1N1)2009 virus. This indicates that it is a Type A virus with H1 and N1 proteins in combination.
• The World Health Organization declared the infection a global pandemic in August 2010

There are three types of seasonal influenza viruses: A, B and C.

• Type A may infect multiple species- Humans, pigs, birds (seasonal flu/epidemics/ pandemics)
• Type B only infects humans (seasonal flu/epidemics)
• Type C may infect humans and pigs (mild respiratory symptoms)

Type A may have sub-types depending upon the combination of two proteins, namely Haemagglutinin (H) and Neuraminidase (N). These proteins may have different numbers:

H: 1 to 17 and N: 1 to 10

The combination of numbers determines the name of the virus. Thus, we have H1N1, H1N2, etc.

Disease transmission:

• Inhalation or ingestion of droplets containing virus from people sneezing or coughing; it is not transmitted by eating cooked pork products.
• People who work with poultry/swine are at increased risk of infection with this influenza virus.

Signs and Symptoms:

• In humans the symptoms of “swine flu” H1N1 virus are similar to those of influenza and of influenza-like illness in general.
• Symptoms include fever, cough, sore throat, body aches, headache, chills, fatigue and sometimes diarrhea and vomiting.
• The most common cause of death is respiratory failure. 
• Fatalities are more likely in young children and the elderly, or previously sick patients like on dialysis, DM, Immunocompromised etc.

Diagnosis: Investigation confirmation by the REAL TIME PCR of nasal and oral secretions. This test in Hyderabad is done by – IPM, Narayanaguda

Guidelines:

Categorization of patients based on risk: Cat A/Cat B/Cat C

Category A

Symptoms/ Signs: Mild fever + sore throat/ cough with/ without bodyache, headache, diarrhea and vomiting.

Treatment: Symptomatic

Oseltamivir: Not required

H1N1 testing: Not required

Monitoring: 24-48 hours by a doctor.

Prevention: Patients should stay at home and avoid mixing with public and high risk members of family.

Category B

This has two sub-categories:

i) In addition to signs and symptoms of Category A, high grade fever and sore throat is present.

May require home isolation and Oseltamivir

H1N1 testing: Not required

ii) In addition to signs and symptoms of Category A, one or more of the following high risk categories is present:

•    Children with mild illness but with predisposing risk factors. 
•    Pregnant women; 
•    Persons aged 65 years or older;  
•    Patients with lung diseases, heart disease, liver disease, kidney disease, blood disorders, diabetes, neurological disorders, cancer and HIV/AIDS; 
•    Patients on long term cortisone therapy.

Treatment: Broad spectrum antibiotics as for Community Acquired Pneumonia

                     Oseltamivir

H1N1 testing: Not required

Prevention: All patients of category B (i) and (ii) should confine themselves at home and avoid mixing with general public; high risk members of their family

Category C

In addition to signs and symptoms of category A and category B, the patient has one or more of the following:

•  Breathlessness, chest pain, drowsiness, fall in blood pressure, sputum mixed with blood, bluish discolouration of nails
• Children with influenza like illness who had a severe disease as manifested by the red flag signs (Somnolence, high and persistent fever, inability to feed well, convulsions, shortness of breath, difficulty in breathing, etc)
• Worsening of underlying chronic conditions.

Treatment: Immediate hospitalization and treatment

H1N1 testing: Required

OSELTAMIVIR (TAMIFLU)

MOA: Oseltamivir inhibits neuraminidase, it must be administered within 48 hours of symptom onset to provide optimal treatment for a selected subgroup of patients.

Dose: 75mg BD

Adverse events: nausea, vomiting, skin reactions and sporadic, transient neuropsychiatric events (self-injury or delirium)

Note: We are not going to discuss here whether tamiflu works or not but do remember Oseltamivir (Tamiflu) isn’t for everyone, and it doesn’t make a difference for most. It probably gives us 1-2 days of symptomatic relief at the cost of antiviral resistance. However, applying this drug to the right at-risk patients may help reduce severity of illness and will hopefully prevent deaths, where even a small therapeutic benefit might provide the right patients an added advantage.

General precautions:

• Frequent Hand Washing
• Covering mouth and nose with tissue paper when coughing/ sneezing
• Avoiding crowded places and those with symptoms of influenza
• Avoiding contact greetings- hugs/ embraces/ kisses/ hand shakes, etc.
• Those with symptoms suggestive of influenza should visit a health care facility at the earliest for early diagnosis and treatment.
• Patients should be provided with three-layered surgical mask in hospitals.

Vaccination

• Not recommended for general public at present.
• Recommended only for Health care workers working in close proximity to influenza patients:
• Those working in the ED/ICU/Isolation wards of hospitals treating influenza cases 
• Those identified for working in screening centres for categorization of patients

Even with appropriate matching with the circulating strains, efficacy of vaccine may be about 70% to 80%. So, vaccine should not give a false sense of security. Considering the risk perspective, the preventive modality of infection prevention and control practices should be strictly followed. The available vaccine takes about 2-3 weeks for development of immunity.

Further Reading:

1. http://www.emlitofnote.com/2014/04/tamiflu-bell-tolls-for-thee.html
2. http://www.who.int/csr/resources/publications/swineflu/h1n1_guidelines_pharmaceutical_mngt.pdf
3. http://www.bmj.com/content/348/bmj.g2545
4. http://www.ncbi.nlm.nih.gov/pubmed/25285542
5. http://www.cdc.gov/flu/professionals/antivirals/summary-clinicians.htm
6. http://mohfw.gov.in/index4.php?lang=1&level=0&linkid=372&lid=3066

Syncope that Kills!

This week we have a case for review,
I came across this 65/M who was brought to the ED with Shortness of breath on exertion and recurrent episodes of syncope over 2 weeks. Otherwise his history was limited. He came in with stable Vital Signs (RR- 20/min) and a normal systemic examination.


What next? Yes, We got the 12 lead ECG for him:

(ACEPs clinical policy on syncope strongly recommends that an ECG be obtained in 
the initial evaluation of patients with syncope)

Interestingly, this was read as:
Computer: Anterior Wall Ischemia
EM Resident: R/O ACS
Cardiology Fellow: Anterior Wall Ischemia

Cardiology and Neurology Consult was asked for, and he got a head CT (though his CNS exam was normal) and got a set of biomarkers from Cardiology (including d-dimer!). And guess what, his troponin was reported normal but d-dimer was elevated! So we ended up getting a CT Pulmonary Angiogram for him.

This guy had bilateral PTE and showed up with HR in 90s, BP 120/70 SpO2 >95% RR 20/min on ambient air. He was taken up for thrombolysis in view of recurrent syncope and possible cerebral hypo perfusion.

This case raised a few questions:
1. Was his presentation concerning enough for this deadly disease?
2. Did we miss anything on the ECG?
3. Why did we do a head CT?
4. Why was he thrombolysed when the teaching is thrombolysis is reserved for massive PE?

Lets go through each one of them:

1. Syncopy and PE
About 10-15% of the patients with pulmonary embolism present with syncope. The textbook history of PE with acute chest pain/SOB and hemoptysis is seen rarely. We need to do a good history asking all the possible risk factors + complete physical exam and then risk stratify them based on clinical decision rules. A common finding which is not documented on the charts is the "calf exam". 
Even after all this, it is not possible to pick every patient with PE! 

In addition to all this we need to watch for other deadly cause of syncope!
Acute Coronary Syndrome
Subarachanoid Hmg
AAA
Aortic Dissection
Ectopic Pregnancy
Arrythmias (Heart Blocks/Brugada/HOCM/WPW/QTc Syndromes/ARVD)
Intra abdominal bleed (Ruptures Spleen/Ovarian Cyst)
Vertebrobasilar TIA
GI Bleed

So, classic presentation of PE is rare. Apart from ACS/Arrythmias we need to think of other deadly differentials of syncope and in the ED - PE should always be in the lost of differentials of syncope!

2. ECG and PE
His ECG was read as Anterior wall Ischemia but then how do we explain the inferior TWI?
ECG is not a great tool to pick up PTE but should be used as only an adjunct to history and physical. Again classic S1Q3T3 is rare.

In fact his ECG showed something more important. There were "Simultaneous TWI in inferior and Precordial Leads" which suggested RV strain. Other ECG changes that may be seen PE are:
Sinus Tachy
RBBB
RAD
P Pulmonale
Atrial Arrythmias
Non Specific changes
NORMAL ECG (ECG in PE can be normal!)

The most important ECG finding for PE, which we should know is probably "Simultaneous TWI in inferior and precordial leads". Pattern of TWI between Acute Pulmonary Embolism (APE) and ACS can be differentiated.

3. Role of head CT while evaluating syncope?

This guy showed up with a normal CNS exam. Is head CT justifiable for such patients with syncope and a normal CNS exam. Well, Strokes generally do not lead to syncope but a rare exception may occur in the setting of global bilateral cerebral ischemia or basilar artery disease affecting the reticular activating system, leading to ischemia that may present as a stroke. However, in the evaluation of a patient with TIA, MRI/MRA (not a CT scan) is a much better choice.

In patients with sudden onset of a severe headache in the setting of a possible syncopal event (with concern for a subarachnoid hemorrhage), doing an emergent head CT on arrival is very reasonable. Patients who present to the ED after a brief syncopal event, without focal neurologic symptoms, headache, altered mental status, or associated trauma, and who are not taking anticoagulants, are unlikely to benefit from a routine head CT.

In 2006, AHA/ACC stated that a neurologic cause of syncope should be considered only if suggested by history or physical examination.

4. When do we thrombolyse PE?

PE can be classified as :

1. Massive: with SBP < 90mm Hg
2. Subamssive: with SBP > 90 but with RV dysfunction/Positive troponin
3. Non Massive: No Hypotension/No RV Dysfunction/Negative Troponin

Thrombolysis is traditionally recommended for PE with hemodynamic compromise, with which most of us would agree. But in the above mentioned case, lytics were administered. Well, this may be justified based on his history of recurrent syncope and elevated Pulmonary artery pressures on ECHO which suggested RV Dysfunction. Thrombolysis always come with risk of bleeding and we need to do shared decision making in such scenarios. There might be a case where patients looks really sick with PTE but holding the blood pressure!

Lets summarise the take home points:

1. Keep PTE always in your list of differentials for syncope.
2. Simultaneous TWI in inferior and precordial leads is PE until proven otherwise.
3. Avoid doing a routine head CT for syncope with unremarkable CNS exam.
4. Thrombolyse massive PE, and for submissive PE go for shared decision making.

For Further Reading:

1. http://www.aapsus.org/articles/42.pdf
2. Soteriades ES, Evans JC, Larson MG, et al. Incidence and prognosis of syncope. N Engl J Med. 2002;347:878-885.
3. Kapoor WN, Karpf M, Maher Y, Miller RA, Levey GS. Syncope of unknown origin. JAMA. 1982;247:2687-2691.
4. Grossman S. Testing in syncope. Intern Emerg Med. 2006;1:135-136.
5. Castelli R, Tarsia P, Tantardini C, et al. Syncope in patients with pulmonary embolism: comparison between patients with syncope as the presenting symptom of pulmonary embolism and patients with pulmonary embolism without syncope. Vasc Med 2003;8:257–61.
6. Courtney DM, Kline JA: Identification of prearrest clinical factors associated with outpatient fatal pulmonary embolism. Acad Emerg Med 8: 1136, 2001.[PMID: 11733290]
7. Ferrari E, et al. The ECG in pulmonary embolism. Chest. 1997;111:537-43
8. Differences in negative T waves among ACS,APE-  Kosuge M, Ebina T, et al – Eur Heart Journal cardiovasc care – 2012
9. Kosuge M, Kimura K, Ishikawa T, et al. Electrocardiographic differentiation between acute pulmonary embolism and acute coronary syndromes on the basis of negative T waves. Am J Cardiol 2007;99(6):817-21.
10. Jankowski K, Kostrubiec M,et al. Electrocardiographic differentiation between acute pulmonary embolism and non-ST elevation acute coronary syndromes at the bedside. Ann Noninvasive Electrocardiol 2010;15:145–50
11. Moderate pulmonary embolism treated with thrombolysis (from the "MOPETT" Trial). Mohsen Sharifi, Curt Bay, Laura Skrocki, Farnoosh Rahimi, Mahshid Mehdipour, “MOPETT” Investigators Am J Cardiol. 2013 January 15; 111(2): 273–277.  Published online 2012 October 24. doi: 10.1016/j.amjcard.2012.09.027
12. Fibrinolysis for patients with intermediate-risk pulmonary embolism. Guy Meyer, Eric Vicaut, Thierry Danays, Giancarlo Agnelli, Cecilia Becattini, Jan Beyer-Westendorf, Erich Bluhmki, Helene Bouvaist, Benjamin Brenner, Francis Couturaud, et al. N Engl J Med. 2014 April 10; 370(15): 1402–1411.  doi: 10.1056/NEJMoa1302097