Furosemide in the Treatment of Acute Pulmonary Edema

Anand Swaminathan, MD, MPH (@EMSwami) is an assistant professor and assistant program director at the NYU/Bellevue Department of Emergency Medicine in New York City.

Case Presentation

A 55-year-old man with a history of congestive heart failure (CHF) and hypertension (HTN) presents with worsening shortness of breath over the last 2 days. Vital signs are:

HR: 132, BP: 210/110, RR: 35, O2Sat: 83%

The patient is in severe respiratory distress. A quick bedside ultrasound reveals numerous B lines. Supplemental oxygen is started and the patients O2 sat increases to 86%. An IV is placed and an EKG reveals sinus tachycardia. A nurse begins to draw up furosemide but you consider other options.


CHF is a common problem in the US with over 5 million patients carrying the diagnosis and 500,000 new diagnoses each year.1 Cardiogenic acute pulmonary edema (APE) occurs when blood backs up into the pulmonary vasculature leading to increased oncotic pressure and leakage of fluid into the alveolar spaces. Essentially, patients are drowning. APE patients suffer from both increased afterload: making it more difficult for the left ventricle to function, and increased preload. As such, the goals of treatment must be directed at both decreasing cardiac filling pressures (preload) and decreasing afterload. Additionally, neurohormonal activation worsens cardiac performance, increases intravascular volume, and increases vascular tone. For decades, the mainstay of treatment in APE has been loop diuretics: mainly furosemide. The continued central role these drugs play highlights a lack of understanding of the underlying pathophysiology of the disease.

Pathophysiology of APE

In the 1940s the cardiorenal model was first put forward as the predominant explanation for APE. It was believed that decreased blood flow to the kidneys led to decreased renal function and fluid retention leading to volume overload. This was the basis for loop diuretics being recommended. However, it was clear that this model was insufficient as it did not explain why the disease progressed or the finding of increased peripheral vasoconstriction from invasive monitoring studies.

The cardiocirculatory model was first put forth in the 1970s. This model argued that peripheral vasoconstriction led to decreased cardiac function and that increased preload and afterload were at the center of the problem. This model explained much of what we see occurring in APE.

Finally, in the 1990s, researchers established the neurohormonal model. In this model neurohormones (norepinephrine, renin, angiotensin, aldosterone) are upregulated in APE. These compounds have vasoactive properties leading to vasoconstriction and increase intravascular volume. Current recommendations for APE treatment are based on the integration of the cardiovascular and the neurohormonal models.


Why Not Loop Diuretics?

The idea of using loop diuretics is based on the idea that patients with vascular congestion are volume overloaded. Let’s look at the actual literature here.

Urban Legend: Patients presenting with APE are volume overloaded.

Zile MR et al. demonstrated that while most patients with APE have increased cardiac filling pressures, most did not have a significant increase from their dry weight on presentation.2

50% of patients had a minimal weight gain (< 2 lbs) on presentation for APE.3

Fallick C et al. argue that it isn’t fluid gain but rather shift in fluids from other compartments, particularly shift from the splanchnic circulation, which is normally very compliant.4

Bottom line: Vascular congestion does note equal volume overload. More than 50% of cases of APE are not associated to significant volume overload.

However, many argue that even in the absence of volume overload, there’s so little downside to giving a dose of furosemide that we should just do it.

Urban Legend: Loop diuretics are not harmful in APE so just give them.

Hoffman JR and Reynolds S showed that patients who got furosemide and or morphine for APE had more complications.5 However, the study was small (n = 57) and had multiple treatment arms.

Francis GS et al. described how administration of furosemide actually led to decreased LV function, increased LV filling pressures, increases in MAP, SVR, plasma renin activity, and plasma norepinephrine levels.6 Essentially, furosemide led to activation of the neurohormonal system instead of turning it off.

Kraus PA et al. demonstrated that PCWP was increased for the first 20 minutes after administration of furosemide.7

Finally, Marik PE et al. summarizes the evidence. Furosemide decreases GFR, activates the renin-angiotensin-aldosterone system, decreases cardiac output, and increases afterload early after administration.8

Bottom Line: Loop diuretics are harmful early in the management of APE.

The first 10 minutes of management for these patients is key so any increases in afterload, preload etc are going to be extremely detrimental.

And let’s not forget that many of our patients with APE have ESRD where no dose of loop diuretic will ever make them diurese. If we aren’t using loop diuretics, what should we be doing?

Non-Invasive Positive Pressure Ventilation (NIPPV)

NIPPV has multifactorial action in APE. It decreases work of breathing, stents open alveoli during the entire respiratory cycle leading to improved gas exchange and, in the case of bilevel NIPPV, decreases afterload.

A number of papers have shown decreased intubation rates and decreased ICU utilization with the use of NIPPV. The most recent study showed a decreased ICU admission from 92% to 38%.9

The key for NIPPV is to start it immediately on presentation to the ED. Even if it doesn’t stave off intubation, it will likely help with preoxygenation.


There are many studies looking at the use of nitroglycerin10, comparing it to fuorsemide11 and looking at high-dose therapy12.

The bottom line is that nitro is recommended for all patients with APE. It reduces preload and at higher doses (> 100 mcg/min) it decreases afterload leading to increased cardiac output and decreased SVR.8


Overall, ACEI have been less studied in APE. The proposed role here is shutting of the renin-angiotensin-aldosterone system and decreasing the neurohormonal drive.

Hamilton RJ et al. found that patients who received sublingual captopril in addition to standard therapy were more comfortable and had a decrease in respiratory failure (not statistically significant).13

Haude M et al. demonstrated in a small study that sublingual captopril improved cardiac index and stroke volume versus nitroglycerin.14

Take-Home Messages

1. There’s minimal or no role for the administration of loop diuretics early in the management of APE. The majority of patients aren’t volume overloaded.

2. Immediate care should focus on NIPPV and administration of nitroglycerin.

3. In patients with ESRD, dialysis is what’s ultimately going to fix the patient.

References / Further Reading

  1. Chung P, Hermann L. Acute Decompensated Heart Failure: Formulating an Evidence Based Approach to Diagnosis and treatment. Mt. Sinai J of Med 2006; 73(2): 506-27.
  2. Zile MR, Bennett TD, St John Sutton M, et al. Transition from chronic compensated to acute decompensated heart failure: pathophysiological insights obtained from continuous
  3. Chaudhry SI, Wang Y, Concato J, Gill TM, Krumholz HM. Patterns of weight change preceding hospitalization for heart failure. Circulation 2007;116:1549 –54.
  4. Fallick C, Sobotka PA, Dunlap ME. Sympathetically mediated changes in capacitance: redistribution of the venous reservoir as a cause of decompensation. Circ Heart Fail 2011; 4: 669-75.
  5. Hoffman JR, Reynolds S. Comparison of nitroglycerin, morphine and furosemide in treatment of presumed pre-hospital pulmonary edema.  Chest 1987; 92: 586-93.
  6. Francis GS, Siegel RM, Goldsmith SR, Olivari MT, Levine B, Cohn JN. Acute vasoconstrictor response to intravenous furosemide in patients with chronic congestive heart failure. Ann Int Med 1985; 103(1): 1-6.
  7. Kraus PA, Lipman J, Becker PJ. Acute preload effects of furosemide.  Chest. 1990; 98: 124-8.
  8. Marik PE, Flemmer M. Narrative review: the management of acute decompensated heart failure. J Intensive Care Med 2012; 27: 343-53.
  9. Liesching T, Nelson DL, Cormier KL, Sucov A, Short K et al. Randomized trial of bilevel versus continuous positive pressure for acute pulmonary edema. J of EM 2014: 46(1): 130-40.
  10. Bussmann W, Schupp D. Effect of sublingual nitroglycerin in emergency treatment of severe pulmonary edema.  Am J Card  1978; 41: 931-936.
  11. Cotter G, Metzkor E, Kaluski E, Faigenberg Z et al. Randomised trial of high-dose isosorbide dinitrate plus low-dose furosemide versus high-dose furosemide plus low-dose isosorbide dinitrate in severe pulmonary oedema. Lancet 1998: 351: 389-93.
  12. Levy P, Compton S, Welch R, Delgado G, Jennett A et al. Treatment of severe decompensated heart failure with high-dose intravenous nitroglycerin: a feasibility and outcome analysis. Ann of EM 2007; 50: 144-52.
  13. Hamilton RJ, Carter WA, Gallagher JE.  Rapid Improvement of acute pulmonary edema with sublingual captopril.  Acad Emerg Med 1996; 3: 205-12.
  14. Haude M, Steffen W, Erbel R, Meyer J. Sublingual administration of captopril versus nitroglycerin in patients with severe congestive heart failure. Intl J Card 1990; 27: 351-9.
  15. Mantuani D, et al. Three-view bedside ultrasound for the differentiation of acute respiratory distress syndrome from cardiogenic pulmonary edema. Am J Emerg Med. 2012 Sep;30(7):1324.
  16. Lobo V, et al. Thoracic ultrasonography. Crit Care Clin. 2014 Jan;30(1):93-117.
Edited by Alex Koyfman, MD

39 thoughts on “Furosemide in the Treatment of Acute Pulmonary Edema”

  1. Excellent review…thank you. Just wanted to add that administration of furosemide may lead to electrolyte imbalances such as hypokalaemia, hyponatraemia, and hypomagnesaemia that may cause or even exacerbate cardiac arrhythmias and increase the risk of sudden cardiac death. The other issue is the development of cardio-renal syndrome and worsening of renal function. I think the key is to outline and recognise Acute cardiogenic pulmonary oedema (fluid redistributed and volume depleted) vs patients with chronic fluid overload and acute exacerbations of CHF where diuresis is still a mainstay.

    1. Excellent points. I agree that redistribution of fluid and not fluid retention is present in APE. Loop diuretics definitely increase risk of these electrolyte issues but unlikely to occur early in management.

      1. Great stuff- agree with Dr. Lahri:

        APE is not chronic insidious onset CHF. The former, “flash” pulmonary edema, should be treated in the ED with nitrates and NIPPV. For the latter, loop diuretics have been a friggin godsend. So CHF patients as a whole do much better since we came up with the “insufficient” 1940s theory.

        But Cards thinks that “flash” pulmonary edema doesn’t happen UNLESS you walk around volume overloaded- which we have data that even after hospitalization for diuresis, many pts. are still fluid positive, as opposed to slightly dry where we’d like them. And using dry weights as your gold standard, I believe relates more to the amount of poop when you weigh them, than a good estimate of fluid status.

        Bottom line, I think your overview is excellent and we should avoid loop diuretics in the initial ED eval. They have no good role in the management of an acutely ill dyspneic pt.

        But after 2 hours has passed and your pt. is so much better that Cards is considering letting them go home, THEN you can make a judgment and, if you think it’s clever, order the Lasix that Cards has been recommending over the phone without looking at your pt. and with side comments about how crazy the ED is for not giving the stuff.


        1. Pik – couldn’t agree more. After the acute process is shut off, sympathetic overdrive is quelled, reassess the patient’s volume status and determine if they can benefit from diuresis. This is similar to using lasix in sepsis. In the ICU, 7-10 days after initial resuscitation, the patient may be liters and liters in the positive. At that point, may be useful to giving them a loop diuretic to get that volume off. But giving a loop diuretic early in management to maintain urine output is crazy. Bottom line, in severe, acute processes, loop diuretics play minimal roll.

          1. I don’t think you can really say that “Bottom Line: Loop diuretics are harmful early in the management of APE.” There’s no definitive findings from any of the studies listed. Moreover, those studies didn’t show ‘harm’ to any of these patients–not that I see at least- no mention of increased hospital length of stay, RRT, other morbidity or any mortality. All theoretical.

  2. Good proposal but I have several observations. 1. “…increased oncotic pressure and leakage of fluid into the alveolar spaces”. Oncotic pressure in pulmonary bed is decreased with increased hydrostatic pressure (Frank Starling Law), therefore there is a fluid leakage into the interstitial and alveolar space. 2. The effect of ACEI can be attributed to a hypotensive effect only? I know not, but this could be argued as well.

    1. I think ACEI play a role in shutting down the neurohormonal axis, particularly RAAS. Articles on ACEI in APE (not many of them) point to only a small reduction in BP with these agents.

  3. Great review Dr. Swaminathan! Love it! BiPAP is amazing and has prevented so many ED intubations. Being at a county facility, while the NTG drip and BiPAP are being setup, I’ve given sublingual NTG (spray/tab; ~400mcg per dose, but only ~150mcg bioavailability) in the meantime, which has worked great in a few anecdotal cases I’ve been a part of. Also, our EM/Cards attendings have really pushed us to give enaliprat (IV version of enalapril; 1.25mg IVP) up front while the pt is on BiPAP and NTG drip is being setup. How has your experience been with sublingual NTG and enaliprat?

    1. I’m a big fan of ACEI in APE. I use them readily in addition to NTG and NIPPV. ACEI are helpful at attenuating the neurohormonal response. Happy to see your EM people are pushing for it’s use.

  4. Pingback: Insuffisance cardiaque aiguë et oedème pulmonaire cardiogénique | thoracotomie
    1. That’s not necessarily true. The trial you are referencing was very poorly done and has significant methodological flaws. Wow relatively larger than the other meta-analyses, it is difficult to have a lot of take away points from a significantly flawed trial.

      There are two small meta-analyses that show short-term mortality benefit.

      Masip et al. JAMA, December 28, 2005: Vol 294, No. 24

      Peter et al. LANCET, April, 2006: Vol 367, No 4.

      And these:

      Thompson et. al. J Emerg Med 2008; 52(3)

      Warner. Prehosp Disas Med 2010; 14(2)

      And the largest ADHF trial ever (retrospective but helpful) ADHERE

      Tallman et al. ACAD EMER MED, April, 2008: Vol 15, No 4.

      1. What are the main methodological flaws and are they worse than small heterogenous meta-analysis?
        The 3CPO study was a pragmatic trial looking at effects of NIV on patients thought to have cardiogenic pulmonary oedema

        1. I am not a research expert, but I did find several things that I
          found troublesome. While there was an intention to treat analysis, 56 out of 347 patients were transitioned to NIV from the standard O2 group. That is 15% of patients in that group, a huge number!! C3PO was also limited by the
          exclusion of severely ill patients. There was a failure of almost 20% of patients to complete their assigned treatments and the above referenced cross contamination between the treatment arms. Lastly, the protocol didn’t use objective criteria for intubation (a major outcome measure), and the reported intubation rate (3%) is much lower than those in other large trials/reviews (~12%).

          BTW. I also found this reference showing a decreased mortality for NIV.

          Meta-analysis: Noninvasive ventilation
          in acute cardiogenic pulmonary edema.

          Cui-Lian Weng, Yun-Tao Zhao,
          Qing-Hua Liu,
          Chang-Jun Fu,
          Feng Sun,
          Yan-Liang Ma,
          Yan-Wen Chen,
          Quan-Ying He

  5. Great stuff Swami. I thought I was the only anti-furosemide Zealot in NYC. (well, I guess that now leaves one of us)

  6. There are SOME patients that benefit from Lasix, however these patients are usually pretty obvious clinically. They are frankly edematous, have a more gradual decompensation, have a history consistent with gradual fluid accumulation, and are by and large relatively normotensive. Check out the Chatti article here for a great review. http://wp.me/p4mPlb-N

  7. We’ve been using furosemide successfully for decades, we’ve all see the patient who pees 2 liters out in the first couple of hours and then looks great. While I agree this data is interesting, we need some real patient-oriented data to show benefit versus harm. There’s no harm data here except a bunch of theoretical arguments, doesn’t seem sufficient reason to abandon a treatment we’ve been using successfully for years. Should be an easy topic to study, if only there were pharmaceutical dollars involved, we’d have much better data.

    1. I think we have used furosemide successfully for decades in patients with CHF exacerbations. The pathophysiology of APE/ADHF is quite different as evidenced by multiple studies (some of which are referenced here). With increased sympathetic tone, flow to the kidneys is minimal. Most patients with ADHF either aren’t volume overloaded or don’t have working kidneys.

      In simple CHF exacerbations, it’s possible furosemide is a fine intervention. In APE/ADHF, more aggressive management directed at the underlying etiology is indicated. Nitro, NIPPV +/- ACEI. Once you’ve turned off the acute pathophysiology, reassess for volume overload and dialyze or add furosemide as needed.

  8. in APE MONA is recommended treatment – Morphini HCl, Oxygen, Nitroglycerin,ASA and NIV, everything else is for considering…

  9. Morphine is outdated and frankly dangerous. The preponderance of evidence (not RDCT unfortunately) demonstrates an association with harm in a number of cardiac disease including ACS and APE/ADHF. Morphine is mechanistically flawed and should be removed from the algorithm completely.

  10. There are a few papers based on the ADHere registry in the US looking at diuretcs and opiates. Opiates are a real dogma problem, independent association with need for intubation.

    Impact of Intravenous Loop Diuretics on Outcomes of Patients Hospitalized with Acute Decompensated Heart F ailure: Insights from the ADHERE Registry.

    Morphine and outcomes in acute decompensated
    heart failure: an ADHERE analysis

    1. great review Swami,
      I believe that randomized data strongly support only the use of NIV in ACPE treatment; however, it seems reasonable to use also nitrates to decrease pre- and after-load, and diuretics only in frankly overloaded edematous patients, generally the 50% of ACPE patients with low EF (systolic ACPE).
      Further help may come from bedside US to identify patients with dilated VCI who may benefit most of diuretics.

  11. Hi,
    I am a junior doctor currently practicing at primary health care that does not have access to NIV or IV GTN drip. It takes us ~30mins or more to transfer the patients to a hospital with cardiac units or 15mins to general hospitals. Recently, we had a similar case as mentioned here except she had a NSTEMI and some cardiac surgery (no documents/ notes in our system) at one of the big hospitals. Not known CHF.
    We do have fursoemide IV and morphine… some of the basic drugs.
    We learnt the basic guidelines to treat pulmonary edema which tells us to use furosemide and I understand the concept of not using furosemide but 1) what do we do in this acute setting without any other available means 2) my patient once she was transferred became well, was started on NIV and GTN and after 80mg of furosemide that we gave did not receive more. She was well and d/c with follow up. Are there certain cases only where the effect of furosemide is negative?

    Thank you

    1. Sharon – thanks for your thoughts. If you don’t have NIV, the answer is going to be high-dose NTG and possibly, intubation if you need to support respiratory. In patients with APE, they are not going to benefit from furosemide as I detailed before. Really the answer is NTG and possibly ACEI if you don’t have NIV.

  12. Thanks Dr. Swami for an awesome review. This issue is a constant battel specially with older physicians who would slam a 120mg of lasix as soon the the APE pt gets through the door. But it makes sense that early on presentation Lasix has no role in the tx of APE.

  13. This is exactly what I did for a patient with ESRD on HD recently in my ED with acute dyspnea. Bilateral pleural effusions and B lines on bedside ultrasound. Gave a tab of Nitro SL while she was on bipap for 30 minutes and did wonders. Hospitalist laughed when I said I hadn’t given lasix, told me that it has direct vasodilatory effect and should be given even if patient anuric. What are your thoughts on this?

  14. Got this comment and wanted to respond:
    This is exactly what I did for a patient with ESRD on HD recently in my ED with acute dyspnea. Bilateral pleural effusions and B lines on bedside ultrasound. Gave a tab of Nitro SL while she was on bipap for 30 minutes and did wonders. Hospitalist laughed when I said I hadn’t given lasix, told me that it has direct vasodilatory effect and should be given even if patient anuric. What are your thoughts on this?

    You did the right thing. Excellent work! I would use this as an opportunity. Show the hospitalist the literature showing no benefit and possibly harm and ask him/her to show you the literature that supports their proposed management. If they show you something that is good, post it up here and we can discuss further. The evidence behind their idea doesn’t exist but they’ll have to discover that on their own.

Leave a Reply