Friday, April 30, 2021

This patient with "NSTEMI" was not allowed to go to the cath lab. Then the ED provider obtained an emergent coronary CT angio. What do you think it showed?

 Submitted by Shakita Crichlow MD, edits by Meyers

A female in her 60s presented with chest pain off and on starting the day before presentation. The chest pain was left sided, pressure-like, intermittent, without aggravating or alleviating factors, and associated with mild shortness of breath. She become worried when she took her blood pressure at home and found it to be 200 systolic, so she decided to come to the ED at that point. 

Here is her initial ECG:

What do you think?

Raw findings:
 - Sinus rhythm
 - STE in leads II, III, and aVF, reaching at least 1.0 mm in III and aVF
 - Large Q wave in lead III
 - STD in aVL and I
 - Subtle STD in V2

Diagnostic of inferoposterior transmural injury, with the most common etiology being Occlusion MI of the RCA until proven otherwise. The normal QRS complex followed by focal inferior STE and reciprocal STD in I and aVL is diagnostic of inferior involvement, while the subtle STD maximal in V2 is diagnostic of the posterior involvement. The Q wave in III is pathologic.

There is at least 1.0 mm of STE in leads III and aVF, objectively and unequivocally. In our recent study  we would have called this STEMI(+) OMI. Interestingly, many providers (including every single one involved in this case) would not call this STEMI even though it meets these objective (and ineffectively insensitive) criteria. Cases like this are why the cardiologist in our study classified EVEN FEWER cases as STEMI than Dr. Smith and I did. I am chronically at a loss for words to respond to the cardiologists who refuse to act according to the same paradigm they simultaneously refuse to give up.

1.0 mm STE in aVF.

The ED provider did not quite recognize the STEMI criteria on this ECG, but clearly was worried about the findings in the inferior leads. They immediately consulted interventional cardiology who refused to take the patient to the cath lab, stating that the patient has only an NSTEMI and does not benefit from emergent management. They asked the provider to call back with the initial troponin.

The initial troponin T (contemporary assay, URL less than 0.01 ng/mL) of course returned undetectable (less than 0.01 ng/mL), as is the case in a very large portion of OMI and STEMI with acute onset symptoms, especially with 4th generation (not high sensitivity 5th generation) assays. 

--Even with hs assays, 10-15% are below the 99th percentile URL at presentation, for both I and T.  
--25-40% are below the 52 ng/L level which the Eur Soc of Cardiology states has an adequate positive predictive value in high risk chest pain patients to diagnose acute MI)

The ED provider had another discussion with the cardiologist, who heard about the negative troponin and suggested that the patient didn't even need to be transferred to the cath lab center. After further discussion, the ED provider ordered a stat coronary CT angiogram. It is unclear to me whether the provider was thinking:

a) "I can see that there is acute high risk ACS, I'm worried about RCA occlusion MI, and if they don't believe it then I'll just prove it to them and keep trying to get her the care she needs."


b) "I don't see clear STEMI criteria, and the first troponin was negative, so maybe this isn't even ACS at all. Maybe I can rule her out for ACS with a normal CT angio."

I am not sure whether the patient had ongoing pain at this time, but a repeat ECG was ordered:
There is still active ongoing injury, but it looks slightly less than the prior ECG. I would not be surprised if the lesion were starting to open just barely.

The coronary CT angio was done within a few hours of her arrival:

Coronary CTA:

"Dominance: the patient is co-dominant.

Left main: no plaque or stenosis.

Left anterior descending: mild multifocal calcific narrowing of the LAD in the proximal and middle thirds with the most severe stenosis estimated at 30-35%.

Left circumflex: mild to moderate multifocal calcific narrowing of the circumflex in the proximal and middle thirds with the most sever stenosis estimated at 40-50%. 

Right coronary: There is minimal punctate calcific narrowing in the proximal third with luminal stenosis estimated at 20-25%. Probable high-grade narrowing in the mid RCA near the acute right marginal branch. Severe motion artifact and blurring at this level limits assessment in the majority of cardiac phases. Degree of maximal coronary stenosis - 70-99% in the mid RCA. 

Interpretation - ACS likely."

Around this time, a second troponin T resulted at 0.73 ng/mL (this is quite high, and rapidly rising, and surely would have peaked higher than 1.0 ng/mL if further troponins had been ordered, which has been repeatedly confirmed in multiple studies as a reasonable retrospective cutoff differentiating OMI peak troponins from non-occlusion MIs).

With this information in hand, the ED provider called the cardiologist back and they now both seemed to understand that the patient likely had an acute occlusion or near occlusion of a major coronary artery (OMI!). 

So the patient was then emergently transferred to the cath lab center, straight to the cath lab.

Another ECG was recorded before transfer:

Definitively reperfusing, with resolution of STE and STD, and with terminal T wave inversion in lead III. The lesion is at least barely open at this time.


RCA culprit lesion with subtotal approximately "98%" lesion with TIMI 3 flow. A stent was deployed with 0% residual stenosis with brisk TIMI 3 flow. 

There was also a 70% LAD lesion which was not deemed a culprit lesion, but was also stented nonetheless.

"IMPRESSION: 65-year-old female patient admitted with non-ST-elevation myocardial infarction. Percutaneous coronary intervention with drug-eluting stent of the right coronary artery..."

The patient did well and was discharged home with a final diagnosis of "NSTEMI."

Learning Points:

This case demonstrates how acute coronary artery occlusion could be better understood if it were evaluated and treated analogously to acute cerebral artery occlusion: using the "Large Vessel Occlusion" (LVO) paradigm. If you present with symptoms of a possible acute occlusion in your MCA, providers universally understand that initial screening tests and exam is insufficient, and the patient must receive emergent imaging to identify LVO. But if you present with possible symptoms of acute occlusion in your RCA, everyone simply looks for millimeters on a piece of paper and calls it a day if they don't find it!

Why is that? It is because CT cerebral angiogram (or MRI, which is also easy) is the diagnostic test of choice for stroke and CT is so easy to do because it does not take a team or special vascular access, and because its only function is diagnosis. If treatment is required, only THEN do patients go to the cath lab.

Perhaps we should be managing MI like stroke: CT angio first, and cath lab activation if CT angio is concerning. One problem is this: imaging the heart with CT is far harder than the brain because of motion. It usually requires no atrial fibrillation and slower heart rate.

Read this post on coronary CT angio:

In this case, a patient with acute RCA OMI was not recognized by ECG (even though it could have been), but instead was recognized by emergent coronary CT angio. Although I and probably all patients would prefer expert level ECG interpretation to recognize OMI as soon as possible and noninvasively, I am very glad that the coronary CT angio was done quickly in this patient! If not, the patient likely would never have been recognized as active RCA occlusion, and would have had greatly delayed reperfusion. In this particular case, the patient reperfused just slightly before the cath. However, many STEMI(-) OMIs will not; furthermore, many reperfused OMIs like this patient will reocclude in the middle of the night and be ignored until the morning when it is too late. 

Ultimately, OMI is not about the ECG. OMI is about the actual, deadly, fixable pathology happening to the human beings we treat: the acute coronary Occlusion Myocardial Infarction. Any method of identifying this pathology (ECG, coronary CT angio, bedside ultrasound) can help the patient get the care they need.

In a normal QRS complex (without any reason for baseline ST segment abnormalities), STE in the inferior leads with reciprocal STD in aVL should be considered inferior OMI until proven otherwise. Similarly, STD maximal in V1-V4 should be considered posterior OMI until proven otherwise.

Wednesday, April 28, 2021

Fever, tachycardia, hypotension, hypoxia and "SVT"

This 40-something presented with hypoxia, BP 60/30, pulse 195, and Temp of 40 C.

He had what appeared to be SVT on the monitor.

Here is his 12-lead:

Narrow complex tachycardia at a rate of 184.

This was interpreted by the computer and the over-reading physician as "SVT"

While it is SVT, the supraventricular part is sinus.  Usually when we say SVT, we are referring to a reentrant rhythm.  This is why I prefer the term PSVT (Paroxysmal SVT), to distinguish sinus or other automatic SVT from re-entrant SVT.

First, when there are generators of sinus tachycardia, such as hypoxia, fever, and hypotension, sinus tachycardia should be strongly suspected.

And, indeed, if you look closely, there are clear P-waves.  This is most easily seen in lead V1, where there is an "up-down" wave just before the QRS.  But you can also see it in many other leads.

See magnified image of Up-down V1 P-wave here:

You can also confirm that the P-wave in V1 is simultaneous with an apparent P-wave in lead II across the bottom:

Other ways to suspect and/or confirm sinus tach vs. SVT:

1. The original heart rate was 195 (measured by the palpated pulse or the continuous monitor) but the heart rate on the ECG is 184.  

Re-entrant rhythms do not change rate gradually.  They are constant, or nearly so (until they break) because the re-entry circuit has a fixed rate.  

The sinus node, on the other hand, is an automatic rhythm, and thus its rate varies gradually and is affected by many many factors, including catecholamines, oxygenation, hemoglobin, volume status, random variation, and many more, and varies with interventions such as oxygen, fluids, and much more many more.  

2. One could apply Lewis leads.  These would exaggerate the P-wave amplitude.  Use the monitor, not the 12-lead ECG, and follow these instructions:

Other cases:

A Relatively Narrow Complex Tachycardia at a Rate of 180.

Heart rate of 230 beats per minute

See here for many uses of Lewis Leads:

Sunday, April 25, 2021

A woman in her 40s with palpitations and chest pressure of unusual etiology

Submitted and written by Magnus Nossen MD from Norway, with some minor edits by Meyers and Smith

A female in her 40s with no known cardiac disease presented to the ED with palpitations and presyncopal episodes recurring over several years, usually lasting 1-5 minutes, sometimes associated with chest discomfort, and increasing in frequency over the past few months. 

Previously she had an echo and 5 days ambulatory ECG performed at a private clinic, both normal. The suspected arrhythmia had evaded capture. She then purchased a smart watch with the possibility for ECG recording. She presented to the emergency room with print outs recorded during palpitations. She was asymptomatic at presentation.


ECG#1: Print out of rhythm strip from her apple watch (25mm/s). As far as one can interpret a single lead, this shows an irregular broad ventricular tachycardia with axis alternation. Duration of this episode was 90 seconds total. From this single rhythm strip differential diagnoses include atrial fibrillation with aberrant conduction/preexcitation, or polymorphic VT (which can be due to long QT, ischemic, or catecholaminergic in variety). Based on the normal beats at the end of the rhythm strip, there is no evidence of long QT. The history was not consistent with that of CPVT. She had not experienced episodes of palpitation/presyncope during physical activity. She was admitted for further work up. 


ECG#2: At rest upon admission: SR, normal interval and axes. Normal repolarization. No sign of preexcitation visible. Normal QTc. (434ms) No Brugada morphology. No epsilon waves. Slight QRS notching inferior leads, but not really consistent with crochetage sign. Interpretation: Normal resting ECG.

Echocardiography revealed a structurally normal heart with normal function. MRI showed a completely normal right and left ventricle with normal systolic function and EF 70%. Stress test was performed adequately without arrhythmia or ischemic findings. The patient was sent for coronary angiogram. This showed patent LMCA, LAD, Cx and RCA. No atherosclerotic changes. Gene test for catecholaminergic polymorphic VT was negative. An ICD was placed. Patient was started on beta blocker discharged home with home monitoring of the ICD. 

After ICD placement and beta-blocker treatment she experienced increasing frequency of paroxysms of palpitations and presyncope. She was admitted again for telemetry and observation. She had several episodes (lasting up to 2 minutes) of palpitation/presyncope before admission and while on telemetry. There was time to record 12 lead ECGs during these episodes. 

ICD Report:

ICD Report: This is from the ICD-report which shows both the atrial channel (top waveform) and the ventricular channel (bottom waveform). The ventricular channel shows irregular ventricular tachycardia with quite a few R-R intervals less than 200ms translating to ventricular rates greater than 300/minute for short duration. There were 14 episodes of VT during one 24 hour period with ventricular rates between 200-283/min. Most episodes were only of a few seconds duration.


ECG # 3: This 12-lead is recorded during one of her episodes. This ECG shows runs of irregular ventricular tachycardia. Slightly varying QRS-morphology resembling RBBB+LAFB. Intermittent cessation with normal sinus rhythm. 


ECG# 4 (limb leads on the left, and precordial leads on the right). This ECG is highly interesting and recorded during another of the patient's episodes. This ECG shows very frequent PVCs with some normally conducted complexes.

However there are striking findings regarding the etiology of the condition! There are several narrow complexes visualized both in the standard leads and in the precordial leads which show “shark-fin” ST-elevation in leads II, III, aVF and V5-V6 with reciprocal ST-segment depression in early precordial leads and lateral leads. The wide complexes show excessively discordant ST segment elevation in II, III aVF and excessively discordant ST-depression in lead I and V1-V2. There is inappropriately concordant ST segment depression in lead I.  As per Smith’s modified Sgarbossa criteria one can sometimes diagnose transmural ischemia from the aberrantly conducted complexes alone –  the narrow complexes of course further support the etiology of intermittent transmural ischemia as the cause of malignant arrhythmia! 

ECG#5 and #5-1:

ECG#5 and ECG#5-1: From the same episode of tachycardia. Similar to ECG number 4, but more tachycardic. 178/minute. My interpretation is that the mechanism of ventricular tachycardia is not re-entry but rather severely ischemic myocardium leading to myocardial irritability triggering PVC w/ foci located in the region of the posterior fascicle of the left bundle branch. 


ECG# 6 This is a rhythm strip from telemetry. The patient had several episodes of tachycardia – as one can observe, there is clear dynamic ST-T pathologic changes with ST-elevation prior to the short episode of polymorphic VT. This again further support the diagnosis and that the observed ST-T changes seen during tachycardia are not type II mediated, even though this was never really suspected as a cause. 


ECG#7: In this ECG one managed to capture initiation of ventricular dysrhythmia. There is some baseline wander – however the ST-elevation is unmistakable, and after being present for some time the ischemia seems to trigger PVCs. 

She was sent for provocation test. Beta blocker discontinued. She was not started on treatment prior as that could mask spasm. 

Angiogram images and videos:

Before provocation

During provocation

Angiography images and videos of the spasm are included above. The provocation test was performed on the LAD. Because the STE distribution in the ECG was RCA domain, they did not do provocation in RCA due to risk of inducing malignant arrhythmia. The spasm test was considered positive (chest pain and ECG changes).

Long acting nitrates and diltiazem initiated. She was discharged home. Home monitoring of ICD has not revealed any significant arrhythmia. Interrogation of ICD after 8 weeks showed no significant arrhythmia. Symptoms of presyncope have disappeared with treatment.

The etiology to the arrhythmia was in this case elusive. The patient had several ECGs recorded during tachycardia. Some with and some without disproportionate ST-segment shift. This case illustrates nicely that knowledge of the Smith modified Sgarbossa criteria is very useful in the clinical setting, here raising strong suspicion of ischemia due to spasm as the cause of arrhythmia. The patient appears to have been given the correct treatment with clinical response.

Learning Points:

Ischemia is one of the most important etiologies of polymorphic VT.

Coronary spasm can be an elusive, transient, and confusing cause of intermittent ischemia and dysrhythmias.

The principles of appropriate discordance (including the modified Sgarbossa criteria) can be applied to almost any wide complex QRS, including even PVCs.

See Dr. Nossen's other(!) case of fascinating coronary spasm here:

Fascinating case of dynamic shark fin morphology - what is going on?

See these cases for using the modified Sgarbossa criteria in the context of PVCs:

Look at the PVCs!!

Look at the PVCs (again)!!

Hyperacute T-waves and Concordant ST Elevation seen in PVCs only

Anterior STEMI and multiform PVCs with Narrow Coupling Interval. When to give beta blockers in acute MI?

Below we reproduce an excellent and detailed comment from Dr. R. Balasubramanian MD from India:

Admission ECG  #2.   There is a slur at J point seen in II,III, AVF and V4,V5 & V6 ( Fig-1).

       Fig-1 ECG #2 - Red arrows point to the J waves in II,III & AVF.


The slur occurs at J point , but there is no J point elevation nor any ST deviation. Their

Infero-lateral  location is concerning , although they are not yet malignant in appearance.

 J point elevation with  horizontal or down-sloping  ST segment are markers for malignant

 ventricular arrhythmias. It is well known that J waves could appear intermittently  and

become  full-fledged only just before or just after ventricular arrhythmias.   In fact,

going through all the seven ECGs , I could spot  subtle as well as typical J waves with classical

 down sloping ST segments as follows:


In ECG  #4, note the J point elevation in the sinus conducted QRS  in AVR,II,AVF & III.

In V4,V5 & V6 also J waves ( subtle) are seen. ( Fig-2) & (Fig-3)

 Fig-2.  ECG #4-  J Point elevation seen in the sinus conducted QRS in AVR,II, AVF & III.


Fig-3  ECG #4-Red Arrows indicating J POINT elevation ( zoomed)


There is progressively increasing elevation of J POINT with down-sloping ST  segment

shown in the zoomed portion of ECG #5 in AVR,II, AVF & III, apparently giving the

shark fin appearance (Fig-4). Also in V5 & V6 similar complexes are seen which look

like  text-book description of malginant J WAVES.(Fig-5)

                    Fig-4    ECG # 5  Showing J WAVES indicated by RED Arrows.




 Fig-5 ECG #5: RED Arrows indicating the malignant J WAVES in V5 & V6 (zoomed)


So,  may I suggest to  include J WAVE SYNDROME in  the final diagnosis of this patient as:


                             J WAVE SYNDROME

                             VASOSPASTIC ANGINA



With regards

Dr.R.Balasubramanian MD

Professor Emeritus of General Medicine


Pondicherry -  INDIA

Wednesday, April 21, 2021

This is really good Prehospital, ED, and Cardiology care. Inferior de Winter's T-waves.

A 50-something y.o. male with history of previous acute MI and stent was shoveling snow.  Shortly thereafter he had the onset of tight chest pain across the front of his chest, without radiation, but associated with diaphoresis and nausea without vomiting. It felt similar to his prior heart attack 9 years ago. He called EMS immediately. 

Here is his first prehospital ECG: 

What do you see?

There is ST depression maximal in V3, and also in inferior leads.  There is minimal T-wave inversion in aVL, by itself a soft sign of inferior MI.  Are the T-waves large?  Does this inferior ST depression with large T-waves represent inferior de Winter's waves?  

So this is an inferior-posterior OMI, very subtle.

Inferior De Winter's waves have been reported before by Sunil Karna

The medics immediately recognized ischemic ST depression. They gave aspirin and sublingual nitroglycerine.  The pain resolved and they recorded another ECG:

The ST depression is gone.  
In fact, one might not have recognized those hyperacute T-waves on the initial ECG until comparing with the next resolved ECG, where the inferior T-waves are now much smaller than they were.  
The T-wave inversion in aVL is also gone.

Medics recognized that this was an acute MI and alerted the ED, even though they did not activate the cath lab.

Thus, when the patient arrived, cardiology had been notified and was in the ED when the patient arrived (it was a weekday during daytime hours).

Another ECG was recorded in the ED:

Completely normal

The initial troponin was below the Limit of Detection.

The ECGs were inspected by the team and the cardiologist, and the cath lab was activated. 

Heparin and ticagrelor were administered. 

Dynamic ST-T changes are a sign of unstable thrombus that is at risk of occluding at any moment!


Culprit is 99% thrombotic stenosis in the proximal LCX at the take off of OM1.  It was opened and stented.

AlsoCo-Culprit is 95% stenosis in the proximal RCA.

LCx: LCX is a Large vessel.

LCX is a co-dominant vessel.

Which wall was affected?  It was the inferior wall.  Both the circumflex and the RCA supply the inferior wall, and they were co-culprits.  There were hyperacute T-waves, and ST depression in inferior leads during ischemia. 

So that initial ECG is manifesting inferior de Winter's T-waves.

Again, these culprits are at risk of re-occlusion at any moment.  But rapid recognition and treatment prevented such an outcome.

Troponin profile (contemporary 4th generation Abbott Architect, URL = 0.030 ng/mL):

Troponin results return approximately 1 hour after the time listed.  Therefore, you can see that, if they had waited for troponins, it would have been 4 more hours before the diagnosis of acute MI would be made.
Had an iSTAT troponin been used, it would have been 7 hours, as its URL is 0.080 ng/mL.

Here is the next day ECG:

No change

Echocardiogram done that day:

The estimated left ventricular ejection fraction is 64%.

There is no left ventricular wall motion abnormality identified.

Almost zero damage done to the myocardium!  By either troponin or echo.

Meyers comment: This is what near-perfect care looks like for Occlusion MI. If everything is done perfectly and immediately, often there would be no such thing as STEMI because it would never get to that stage, but instead would be prevented from happening altogether, as in this case. 

Monday, April 19, 2021

This case was flagged as a false positive cath lab activation. Why? And do you agree?

An 80 - something with h/o hypertension and hyperlipidemia complained of chest pain off and on all day, then recurring acutely.  He called 911. 

Prehospital ECG

What do you think?

The computer called it ***STEMI*** and the medics activated the cath lab.  

Smith: The computer was correct.  STE in leads I and aVL which is not only diagnostic of OMI, but has 1 mm in 2 consecutive leads (STEMI criteria), with a deep Q-wave in aVL.  There is also STE in V2 with a deep QS-wave.  There is STE in V1 and STD in V5, V6, suggesting an LAD occlusion proximal to the septal perforator with septal injury and reciprocal STD in lateral leads.  Most importantly in confirming OMI, there are hyperacute T-waves in I and aVL and V2-V4.  

The QS-wave in V2 does not appear to be "old MI with persistent STE" (LV aneurysm) because there is a large T-wave: the T/QRS ratio = 4/7 = 57% (greater than 36% in any one of leads V1-V4 is indicative of acute -- vs. old -- MI).

The ECG is diagnostic of acute proximal LAD occlusion or first diagonal occlusion.

Case progression  

Initially the patient rated the pain as 6/10 for severity, but pain was down to 1/10 after prehospital aspirin and NTG, after which he arrived and had this ED ECG recorded:

STE in I and aVL is almost resolved, and hyperacute T-waves are all resolved.  There was an old ECG for comparison:

Previous ECG:

 Old Q-wave in aVL.  Otherwise normal.

Angiogram: Conclusions/Summary


"Indication: Chest pain with dynamic EKG changes concerning for ACS"

--CAD with moderate stenosis of ostial left main.

--CAD with long segment of serial stenosis of proximal to mid LAD.

--Successful PCI of proximal to mid LAD with placement of 3.5 x 38 and 3.5 x

LAD: Large caliber vessel.

There is a long segment of serial 50-80% stenosis noted in the mid portion of the vessel. The LAD is severely tortuous and there is large caliber diagonal after an acute bend in the mid vessel. The distal and apical segments are without significant stenosis.


Lesion on Mid LAD was stented in a complicated procedure.



Estimated LV ejection fraction-lower limits of normal; 54%.

Regional wall motion abnormality-septum, hypokinesis.

Left ventricular hypertrophy, concentric-mild.

This is the interesting note, and the reason this was flagged as a false positive activation:

"This is an 8X-year-old gentleman with past medical history significant for hypertension and hyperlipidemia who presents as a prehospital Cath Lab activation for anterior ST elevation myocardial infarction.  Subsequent prehospital EKG showed dynamic changes with improvement in ST elevation following aspirin and nitroglycerin.  Inferior ST depression and sub-millimeter ST elevation in the high lateral leads associated with small Q waves persisted. 

"His EKG on arrival did not demonstrate ST elevation.  Due to his persistent chest pain and high risk EKG features he was taken emergently to the cardiac catheterization lab where he was found to have moderate left main disease and 90% proximal LAD status post PCI.

"Ultimately this is not a true STEMI, but rather ACS/high risk NSTEMI based on angiographic features which did not show an acute thrombotic lesion, but rather what appears to be more long-standing disease. He remained hemodynamically stable during the case.  He was chest pain-free at its conclusion.

Diagnosis: NSTEMI 

But this was indeed a STEMI (+) OMI!!

A STEMI that resolves is still a STEMI.  It is not NSTEMI.  It called a transient STEMI.  

In many cases, the thrombus can lyse spontaneously and not show on the angiogram.

In my view, a transient STEMI still needs emergent cath lab activation.  Not everyone agrees.

See this post:

Timing of revascularization in patients with transient STEMI: a randomized clinical trial

In this trial, 142 patients with transient STEMI were randomized to emergent vs. next day angiogram with PCI.  MRI measure infarct size was the same in both.  All patients received aspirin, a P2Y12 inhibitor, and an anticoagulant.

However, and this is a big however4 patients in the delayed group had recurrent ischemia and needed to go emergently to the cath lab.

Thus, as I have frequently promoted, if cath is delayed for patients with transient STEMI, one must have continuous 12-lead ST segment monitoring.

Some other interesting cases of Transient STEMI:

Transient ST elevation, rules out for MI, what is it?

Spontaneous Reperfusion and Re-occlusion - My Bad Thinking Contributes to a Death

Recommended Resources