Tuesday, December 29, 2015

Lazarus Phenomenon: Back from the Dead (a.k.a. The 'Real' Walking Dead)

What the Lazarus?!

  • Lazarus Phenomenon a.k.a. Lazarus Syndrome is not to be mistaken for Lazarus Sign or the terrible film Lazarus Effect.
  • It is defined as the delayed Return of Spontaneous Circulation (ROSC) after stopping resuscitative efforts. Usually ROSC in Lazarus Phenomenon occurs within 10 minutes of cessation of CPR. 
  • While it was first documented in 1982, and is considered to be rare, it is also considered to be underreported.

Case History:

  • A geriatric patient with history significant for C2 Odontoid Fracture with quadriplegia, recurrent VRE positive UTIs with Sepsis, and ESRD on hemodialysis who was obtunded on admission. The patient was noted to have a potassium level over five on admission that was persistent, despite HD.
  • On the day of the code, her K+ level was 6.0 prior to temporizing measures. Once the code was called, it was run for 25 minutes with 6 rounds of epinephrine, 2g CaCl, 2g NaHCO3, and 450mg MgSO4 being delivered over that time, as well as several shocks for Ventricular Fibrillation. The patient was simultaneously intubated. At that point, a decision was made with the family present, including the POA, to discontinue ACLS and CPR and allow the family to enter the room and grieve.
  • Monitors were shut off approximately 2 minutes after calling the code with only PEA and no pulses or respirations on palpation, doppler, or auscultation and absent brainstem reflexes. After approximately 10 minutes, the family reported a sudden, spontaneous breath.
  • Upon re-entering the room, the patient was exhibiting agonal breathing with a strong carotid pulse. Bedside US by the MICU attending was performed and EF was 50%, so the patient The MICU was contacted and the patient was transferred.

Proposed mechanism:

  • AUTO-PEEP (Who?!) ... Auto-PEEP, man ... - The proposed mechanism (with some room for hand waving and jiggery-pokery) is the buildup of intrathoracic pressure, particularly in COPD patients, that can result in pericardial tamponade. This can happen acutely during CPR when there is rapid ventilation of the lungs without adequate time for exhalation. Chest compressions are meant to compress the heart and have the simultaneous effect of increasing intrathoracic pressure, but that effect is intended to be transient, as the "recoil" between compressions is essential time for the ventricle chambers to refill with blood to be pumped out during the next compression.
  • The combination of chest compressions and Auto-PEEP could be the key factor where there is too much intrathoracic pressure buildup for the heart to refill between beats and thus to get adequate blood out on it's own to perfuse the body.
  • There's also the matter of chest wall elasticity itself, and as long as we're postulating, patients with decreased chest wall elasticity would also be prone to retaining intrathoracic pressure buildup between beats, further promoting tamponade and preventing ROSC.
  • In such cases, time to allow the auto-PEEP and built-up external pressure from CPR to resolve could be the key to delayed ROSC.

Other proposed mechanisms:

  • Hyperkalemia - It is thought to leave the heart retractile for a longer period of time and potentiate the peripheral blood vessels to continue moving blood slowly in the absence of a heart beat.
  • Delayed effects of drugs - In patients with drugs delivered through a peripheral IV (rather than a central line), the combination of impaired venous return from dynamic hyperinflation (see Auto-PEEP above) and local vasoconstriction caused by drugs such as Epinephrine, could delay delivery of these medications to the central circulation and cardiac vasculature
  • Myocardial Ischemia with "Stunning" - Caused by Hypovolemia.
  • Transient Asystole - Transient period of asystole or PEA after a defibrillatory shock of Ventricular Fibrillation uses up more ATP in myocardial cells than primary asystole or PEA, requiring more oxygen or leading to more lactic acid buildup to replenish these stores.
  • Pseudo-PEA - Heart beats cause too weak a pulse to palpate. This problem can be solved with doppler or bedside echocardiography. In our case, doppler was unable to reveal a pulse in multiple sites.

Why it fits our case?

  • Auto-PEEP - For reasons described above ...
  • 25 minute code! - The average time of the documented cases was 27 minutes of CPR
  • Hyperkalemia - While there are many hyperkalemic patients with ESRD who code, the constellation of factors seemed to be in her favor.
  • Delayed effects of drugs - The patient did not have a CVC in place until after being moved to the MICU.
  • Delayed ROSC - The mean time for ROSC in these cases was noted to be ~7-8 minutes, while it took ~10 minutes in our case. The true time of ROSC could have been anywhere between 2-10 minutes, as leads were disconnected ~2 minutes after recorded time of death, though the first observed breath was at approximately 10 minutes.

What's next?

  • Global Hypoxia/Anoxic Brain Injury! While the patient has ROSC, she's had no blood flow to her cerebral vasculature (or any peripheral vasculature or her heart and lungs, for that matter).
  • She was noted to have "extensive ischemic change" on CT, as well as similar findings on MRI. While her neurological function was limited to begin with, she lost brainstem reflexes and evidence of consciousness. She withdrew to noxious stimuli, but was never arousable. Her EEG by Neurology also showed electrocerebral inactivity indicating that much of the patient's cortical function was lost.
  • This fits the course described by the primary literature, where only 35% of patients with delayed ROSC have been able to return home neurologically intact.

Dessert (for Harsha):

  • Most things in medicine, and science in general, are named after ourselves, people we love, people we hate, the classics, and religious texts.
  • This disorder has it's name rooted in a major biblical reference to the resurrection (i.e. ROSC) of Lazarus of Bethany by Jesus Christ four days after his death and burial, described in the Gospel of John.

Monday, December 28, 2015

K 8.7 mmol/l ! You want to know what the EKG shows??


  • Women in her 60's presents with septic shock from HCAP, currently bacteremic. She was oliguric. On admission her Serum Potassium was 8.7 mmol/l. Prior to admission medications included PO Bactrim for PCP prophylaxis. 
  • Focus of this article is to display the EKG changes associated with hyperkalemia in this patient. 

EKG at 1am: 

Initial EKG with K 8.7 mmol/l showed:

  1. Almost sine wave pattern
  2. Wide complex tachycardia 

We treated the patient with IV calcium gluconate, Normal saline boluses, IV Furosemide, Albuterol nebulizer, Sodium bicarbonate drip, Insulin Dextrose.


EKG at 2am: 

One hour post treatment, K was 7.1mmol/l. A repeat EKG showed:

  1. 1st degree AV block
  2. Prolonged QRS interval
  3. QTc prolongation
  4. Resolution of sine wave pattern.

We repeated the above mentioned treatments, consulted renal (just to feel safe).


EKG at 3am: 

One hour later, K was 6 mmol/l. A repeat EKG showed:

  1. Non specific ST changes. Otherwise normal EKG
  2. Resolution of AV block. 
  3. QRS no longer prolonged. 

Eventually her urine output improved with IV fluids and lasix. K level normalized in 6hrs. Did not need dialysis.



  • Hold bactrim as it can cause hyperkalemia due to blockade of the collecting tubule sodium channel by trimethoprim (an action similar to that induced by the potassium-sparing diuretic amiloride); this is most common in HIV-infected patients who are treated with high doses, but normal doses can produce a modest elevation in the plasma potassium concentration in non-HIV-infected subjects
  • Not every hyperkalemia patient has tented T waves on EKG. Peaked T waves are only one of the many EKG manifestations of hyperkalemia. 



As a general rule, mentioned below are the EKG findings associated with various stages of hyperkalemia:

  1. K >5.5 mmol/L is associated with repolarization abnormalities
      1. Peaked T waves (usually the earliest sign of hyperkalaemia)
  2. K >6.5 mmol/L is associated with progressive paralysis of the atria:
      1. P wave widens and flattens
      2. PR segment lengthens
      3. P waves eventually disappear
  3. K >7.0 mmol/L is associated with conduction abnormalities and bradycardia:
      1. Prolonged QRS interval with bizarre QRS morphology
      2. High-grade AV block with slow junctional and ventricular escape rhythms
      3. Any kind of conduction block (bundle branch blocks, fascicular blocks)
      4. Sinus bradycardia or slow AFDevelopment of a sine wave appearance (a pre-terminal rhythm)
  4. K >9.0 mmol/L causes cardiac arrest due to:
      1. Asystole
      2. Ventricular fibrillation
      3. PEA with bizarre, wide complex rhythm

1. www.lifeinthefastlane.com
2. UpToDate

Thursday, December 24, 2015

"That is a huge PE!!" Massive versus submassive Pulmonary embolism? tPA versus NO tPA.


  • Women in her 60's with PMH of Breast cancer stage 1A currently in remission presented with 1 day history of dyspnea on exertion, dizziness on exertion. She was functionally active prior to the onset of these symptoms. No Hypoxemia and no hemodynamic compromise.
  • CT PE study revealed extensive bilateral PE with occlusive segmental and sub-segmental emboli (Saddle embolus). Troponin was elevated at 0.9, BNP 600's. (no known cardiac history). STAT ECHO showed mildly enlarged RV, with normal RV and LV function. BLE venous dopplers revealed extensive DVT in both lower extremities.
  • This is probably the largest PE that I have seen on a CT scan so far.


  • What would you do now?
  • Does this patient have a massive or submassive PE?
  • Does she have right heart strain? What is right heart strain?
  • Do you give tPA or heparin only?
  • Is there a role for Interventional radiology in this case? 


  • This is a stable patient (normal hemodynamics, no hypoxemia) with evidence of right heart strain.
  • The patient has an acute unprovoked submassive / intermediate PE
  • No indication for tPA in this patient.
  • Patient received heparin only.
  • No role for IR in this case.


* Difference between Massive vs Sub-massive PE?

  • Massive PE: Acute PE with sustained hypotension (SBP <90 mm Hg for at least 15 minutes or requiring inotropic support, not due to a cause other than PE, such as arrhythmia, hypovolemia, sepsis, or LV dysfunction), pulselessness, or persistent profound bradycardia (heart rate <40 bpm with signs or symptoms of shock).
  • Submassive PE: Acute PE without systemic hypotension (SBP >90 mm Hg) but with either RV dysfunction or myocardial necrosis.
    • RV dysfunction means the presence of at least 1 of the following:
      • RV dilation (apical 4-chamber RV diameter divided by LV diameter >0.9) or RV systolic dysfunction on echocardiography
      • RV dilation (4-chamber RV diameter divided by LV diameter >0.9) on CT
      • Elevation of BNP (>90 pg/mL)
      • Elevation of N-terminal pro-BNP (>500 pg/mL); or
      • Electrocardiographic changes (new complete or incomplete right bundle-branch block, anteroseptal ST elevation or depression, or anteroseptal T-wave inversion)
    • Myocardial necrosis is defined as either of the following:
      • Elevation of troponin I (>0.4 ng/mL) or
      • Elevation of troponin T (>0.1 ng/mL)
  • Size of PE is not the only variable that dictates whether PE is massive vs submassive. It is a constellation of the above mentioned findings that dictate if a PE is massive vs submassive.

* Who gets tPA?

  1. Patients with massive PE (above mentioned criteria) and no contraindications to tPA.
  2. For most patients with acute PE who do not have hemodynamic compromise, tPA is NOT recommended. Treat with heparin only. Acute PE with right ventricle (RV) dysfunction constitutes a spectrum of severity and more data are needed before thrombolytics can be routinely administered in this population of patients. However, thrombolysis may be considered on a case-by-case basis when the benefits are assessed by the clinician to outweigh the risk of hemorrhage.

* Do we consult interventional radiology?

  1. In unstable PE patients, once it is decided that thrombolytic therapy is warranted, evidence suggests that tPA be administered by a peripheral venous catheter, rather than a pulmonary arterial catheter. So no absolute indication to call IR right away.
  2. In sub-massive PE patients, if the clinician decides to give tPA, the optimal method of administering tPA is unknown. If tPA is to be administered in acute PE with RV dysfunction, a discussion between IR and an intensivist is required to decide the preferable route based on the local expertise.

Wednesday, December 23, 2015

Mixed Alkalosis in an ESRD Patient (or: How I Learned to Stop Worrying and Love Hemodialysis)

This one kind of bridges the gap between Pulmonary and Renal disorders in the ICU.

Case History:

  • A relatively young woman with ESRD for more than 10 years, presenting awake and alert, breathing well on 2L O2 (her home dose) and a pH of 7.6
  • The patient had gone for hemodialysis as per her usual schedule earlier in the day, but was noted to be behaving confused afterwards by her bus driver and was taken to the ED as a precaution
  • The initial ABG looked more like Metabolic Alkalosis, but over the course of the night, her pH increased to 7.73 as her respiratory alkalosis worsened

Why it’s weird?

  • We don’t typically see alkalosis in ESRD patients! Acidosis is far more common.
  • Instead of compensating her Metabolic Alkalosis with her respirations, she developed a concurrent Respiratory Alkalosis overnight
  • Our Nephrologist’s initial response was, “Curious” …

The differential:

  • Hypokalemia - not exactly the bread and butter of ESRD patients, but we initially wondered if too much potassium was dialyzed, resulting in an Intracellular Shift in hydrogen ions to compensate, but alas, her K was just over 4 on admission.
  • Gastrointestinal losses - these tend to present in patients as emesis or diarrhea. The patient had not had any of these symptoms.
  • Renal losses - these occur through loop and thiazide diuretic use or a variety of inheritable ion transport syndromes (Bartter, Gitelman, and Pendred). Since the patient didn’t make urine, these renal losses were not really an issue.
  • => Mineralocorticoid excess - this is a subset of “renal losses” because it stimulates the mechanisms promoting proton excretion in the urine (collecting duct/tubule proton pump, ENaC channels, and the sodium-potassium pump). The patient had changed physicians and had missed getting refills of some of her medications. One of her antihypertensive medications - Minoxidil - has been shown to increase Plasma Aldosterone Clearance, so missing doses of this could theoretically cause a rebound aldosterone excess, but this was considered a low-likelihood scenario
  • Contraction Alkalosis - this typically occurs with large NaCl losses without concurrent HCO3 clearance. Since the patient went to hemodialysis prior to admission, this could be a possible cause
  • Post-Hypercapnia - not really a consideration as the patient had no reason to have a rapid drop in pCO2 and she doesn’t make urine (which is part of the compensatory mechanism)
  • Weird ingestions - these can include Sodium Polystyrene Sulfonate (commonly known as Kayexalate), antacids, and milk, all of which she denied taking
  • => As an aside on this one, the overconsumption of milk is classically part of the “Milk Alkali Syndrome” with presentation of the triad of Hypercalcemia, Metabolic Alkalosis, and Acute Kidney Injury. The colloquially named “Modern Milk Alkali Syndrome” is associated with antacid use, usually Calcium Carbonate. In renal failure patients, Secondary Hyperparathyroidism is often managed with Calcium Carbonate, to act as a Phosphate-binding agent. Our patient was used a sibling drug called Phoslo (Calcium Acetate). This diagnosis would be hard to prove because two of the classic signs were not present, i.e. the patient already had renal failure so AKI would go unnoticed and her calcium levels were normal, but the calculations were likely skewed by her use of the calcium-chelating Sensipar (Cinacalcet).

The workup:

  • LYTES!! - Na, K, Mg, Cl, HCO3, Ca, PO4
  • Trend your ABG
  • EKG
  • Saline challenge and Urine chloride - this is recommended as part of the initial workup from UpToDate to our friendly purple Pocket Medicine books. Unfortunately, the whole ‘not peeing’ component of her ESRD kind of threw a wrench in this plan. With the respiratory status at baseline, we did eventually give normal saline at a slow rate (50mL/hr) with the hope that the patient could replace some of the bicarbonate with chloride and compensate with some respiratory acidosis
  • Screening for classic abuse and overdose (urine drug screen, urinalysis with microscopic analysis, salicylate level, acetaminophen level) - while we couldn’t test her nonexistent urine, her other medication levels were normal
  • Endocrine testing - Thyroid function tests, Cortisol level

What we did:

  • Slow infusion rate of 0.9% NaCl (50mL/hr) - for chloride replacement
  • IV PPI (Protonix injections) - to slow
  • Holding her Phoslo (Calcium Acetate)
  • Ativan (Lorazepam) low dose 0.5mg Q8H scheduled to slow the respiratory rate and reduce CO2 loss through respiration
  • As her alkalosis worsened, we repeated Hemodialysis with a low-bicarbonate solution and partial fluid replacement with 0.9% NaCl

What we had left in our arsenal:

  • Hydrochloric Acid infusion … not a solution for the basic layman (pun intended) or the faint of heart!
  • Ammonium Chloride infusion
  • Continuous Renal Replacement Therapy (CRRT)


  • The hemodialysis restored her pH to 7.33
  • At no point, even with a pH of 7.73, did the patient show any physical signs except for slightly slowed mentation
  • With no clear diagnosis, it’s possible that this will recur, but seeing as the patient is dialyzed thrice-weekly, the metabolic alkalosis, like hyperkalemia for most ESRD patients, might be managed by exogenous clearance

What we think happened/Take-home points:

  • Contraction Alkalosis from hemodialysis, but the patient’s symptoms have been going on for a few days, so it’s possible her HD formulation changed recently
  • Non-classic Modern Milk Alkali Syndrome with surreptitious or overuse use of her Calcium Acetate
  • If this metabolic alkalosis recurs, and given her history of early-onset ESRD likely secondary to hypertension, she would be a good candidate for additional workup for Hyperaldosteronism
  • Metabolic Alkalosis DDx - Hypokalemia, GI losses, Renal Losses, Mineralocorticoid Excess, and Milk Alkali Syndrome or other ingestions (e.g. Antacids + Kayexalate)

More common presentations:

  • ESRD patient post-HD presenting with metabolic alkalosis and hypokalemia. Monitor for EKG changes and cardiopulmonary symptoms and delicately replace potassium
  • Severe metabolic alkalosis in a patient who pees? Saline challenge, monitor Urine chlorine, replace chloride and potassium as needed, and know the diuretics potassium-sparing diuretics (Amiloride and mineralocorticoid receptor antagonists such as Spironolactone) as well as Diamox (Acetazolamide) could be used in edematous states, and that HD/CRRT or HCl might be necessary in severe or refractory cases.

  1. Sabatine, Marc S. 2014. Pocket Medicine, Fifth Edition.

Wednesday, December 9, 2015

Hmmm. Never seen this color in urine bag !


  • We were rounding on a middle aged male patient, h/o pelvic trauma (has a chronic suprapubic catheter). The catheter is connected to a urine bag, picture of it is pasted below : 
  • PURPLE COLOR URINE in the bag. What is this? 
  • I have never seen this before. I point this out to my intern, and she goes, " Yea! I know this; it is PURPLE URINE BAG SYNDROME." 

Q. What is purple urine bag syndrome? 
  • Rare syndrome where purple discoloration of urine occurs in people with urinary catheters and co-existent urinary tract infection.
  • The purple color of the urine is due to metabolic products of biochemical reactions formed by bacterial enzymes in the urine. 
  • Geeky stuff, try not to read it : Tryptophan in the diet is metabolized by bacteria in the gastrointestinal tract to produce indole. Indole is absorbed into the blood by the intestine and passes to the liver. There, indole is converted to indoxyl sulfate. Indoxyl sulfate is excreted in the urine. In purple urine bag syndrome, bacteria that colonize the urinary catheter convert the indoxyl sulfate to indirubin and indigo; and hence purple discoloration of the urine, collecting bag, and tubing. 
  • Bacteria capable of producing these enzymes include Providencia stuartii, Providencia rettgeri, Klebsiella pneumoniae, Proteus mirabilis, Escherichia coli, Morganella morganii, and Pseudomonas aeruginosa. 
  • People with purple urine bag syndrome usually do not complain of any symptoms. Purple discoloration of urine is often the only finding, frequently noted by caregivers. It is usually considered a benign condition, although in the setting of recurrent or chronic urinary tract infection, it may be associated with drug-resistant bacteria
  • Medical management of purple urine bag syndrome does not require any special treatment apart from changing the catheter and administering appropriate antimicrobial therapy to treat the underlying bacterial infection.

Monday, December 7, 2015

3am in the ICU, a puzzled RN approaches "Hey Doc! This patient is breathing funny on the monitor !!"


  • Middle aged female was admitted to ICU for Metabolic encephalopathy. She has a history of ESRD on HD, missed last 2 sessions of dialysis. Also has dCHF from uncontrolled HTN. 
  • Working differentials were uremic encephalopathy; polysubstance overdose; septic encephalopathy; hypertensive encephalopathy (systolic BP on admisssion was 220mm Hg). 
  • BP was treated with nicardipine drip (hypertensive emergency), later nephrology team dialyzed her. Overnight, while patient was asleep, the monitor showed the following breathing pattern: 


  • This pattern was present for more than an hour, making the RN uncomfortable. Oxygen saturation (SaO2) was normal. 
  • This patient seemed to have Cheyne Stokes Variant (CSV). No apneic spells were noted. After noting this pattern, we have started CPAP. 4 hours after initiation of CPAP, the breathing pattern returned to normal. 
  • CT head did not show any bleed. However, work up for PRES syndrome in underway. Her dCHF and acid base imbalance might have contributed to this breathing pattern. There is also association of central sleep apnea with this breathing pattern. So further outpatient work up with polysomnography should be pursued. 



  • CSV is associated commonly with CNS pathology (strokes, bleeds) and heart failure. 
  • In CSV there is hypopnea but no apneic spells, in contrast to Cheyne Stokes breathing where apneic spells replace the hypopneic spells.  
  • Cheyne-Stokes Breathing : Cyclic crescendo-decrescendo respiratory effort and airflow during wakefulness or sleep, without upper airway obstruction. 
  • When the decrescendo effort is accompanied by apnea during sleep, it is considered a type of central sleep apnea syndrome. 


  • Related to PaCO2 variations in patients with heart failure, Central sleep apnea, neurologic disease, sedation, normal sleep, acid-base disturbances, prematurity, and altitude acclimatization. 


  • These patient must get an overnight polysomnogram = gold standard diagnostic test. 


  1. Treatment of the underlying cause = Optimizing heart failure etc 
  2. Nocturnal continuous positive airway pressure (CPAP), There is sufficient data to show that CPAP may improve cardiac function, blood pressure, exercise capacity, and quality of life in these patients. 
  3. Supplemental oxygen, or adaptive servoventilation (ASV)


  • Unfortunately, none this time ;)

Monday, November 16, 2015

Inhaled steroids in addition to systemic steroids in asthma exacerbations ? WHAT !!

Another pulmonary topic ! Please don't get bored of pulm now. Two more weeks of pulm and then I'll be in MICU.


  • We were consulted for a basic bread and butter Asthma exacerbation in a middle aged obese African American women. Patient had a history of intermittent asthma at home, taking only prn albuterol. 
  • We were doing all the usual right things with nebulizers, IV solumedrol, antibiotics etc. The only difference this time was, I had a new question that didn't occur to me for the last 2.5 yrs of residency. Blame me!
  • Prior to admission, the patient was not on inhaled corticosteroid (ICS). She is now on IV solumedrol 80mg TID for the acute exacerbation.




  • For the last 2.5 yrs, I never added an ICS when the patient was being treated with SCS. Makes total sense right ?! Why would the patient need additional ICS when she is getting SCS. I never tried to look up evidence till today.
  • Well, turns out that adding ICS to SCS showed improved outcomes in asthma exacerbations.
  • There was a well designed RCT by Rowe et al, published in JAMA that looked at this exact clinical question. Adding ICS to SCS during an exacerbation reduced the relapse rates, improved the quality of life, reduced the frequency of rescue albuterol use.

Few limitations that might question the clinical applicability of these results are:

  • There are no other studies that looked at this clinical question. These results need to be replicated in future studies to confidently apply the conclusions in clinical practice.
  • This study was limited to Asthma patients who were only on prn albuterol prior to admission (patients on ICS or SCS prior to admission were excluded). So results may not be equated to other patient populations. 
  • Patients in this study were not treated inpatient. They were discharged from ER on ICS + SCS. So, we aren't sure if this would be applicable to inpatient population. 


  • Starting today, I will definitely add ICS to SCS when a patient is admitted for asthma exacerbation (specially those who weren't on ICS prior to admission). 
  • They should be discharged on ICS along with the SCS burst/taper. At outpatient follow-up, based on the patients clinical status, we could step-down from ICS if necessary. 
  • I guess it may also be ok not to start an ICS - due to the above mentioned limitations. 
  • Well, there is one good study :) Change has to begin somewhere! 


  • The asthma exacerbation patients we usually admit are already on ICS prior to admission. 
  • When these patients come in for an exacerbation, no one knows whether continuing ICS in addition to SCS would improve outcomes. 
  • Pulmonary team attendings at our institution go either ways (Few are ok with holding the ICS while patients are on SCS, while others are very particular about continuing ICS in addition to SCS). 
  • Even in this patient group, my personal opinion (extremely weak evidence, Probably Grade B/C recommendation :) is to continue the ICS in addition to SCS. There is no right / wrong answer. 
Pitch in your thoughts.

  1. Rowe BH1, Bota GW, Fabris L, et al. Inhaled budesonide in addition to oral corticosteroids to prevent asthma relapse following discharge from the emergency department: a randomized controlled trial. JAMA. 1999 Jun 9;281(22):2119-26.
  2. ACP J Club. 2000 Jan-Feb;132(1):13.