Also known as: Coughing up blood

1. Description of the problem


Expectoration of blood originating from below the level of the vocal cords. Distinguish from hematemesis and epistaxis. When massive, is frightening to both patient and physician. High mortality in some reports (22% to 78%).

Continue Reading

Clinical features

Ranges from expectoration of blood-tinged sputum, to asphyxiating, or rarely, exsanguinating hemorrhage. Most episodes, both minor and major, may resolve spontaneously, but only temporarily, with a high risk of recurrence, especially with more severe bleeding. Definition of major hemoptysis varies from 100 mL to 1000 mL expectorated in 24 hours. Quantification by patient or physician is unreliable: practical definition is “coughing up full mouthfuls of blood” and screens for the patient at potential risk of clotting off the proximal tracheobronchial tree (100-200 mL).

Key management points

Massive hemoptysis can be explosive, and the mnemonic “RACE” used for response to fires can be applied to this condition.

Rescue the patient and the uninvolved lung: secure the airway, oxygenation, support hemodynamics and, of critical importance, prevent the uninvolved lung from being drowned in blood. As little as 400 mL of blood in the alveoli is sufficient to severely compromise oxygenation.

Alert support services: Interventional Radiology, Pulmonary and Thoracic surgery — a multidisciplinary approach, instituted immediately, is key to ensuring the best outcome for the patient.

Contain the bleeding source: Not uncommonly this involves some form of tamponade of the bronchial orifice exuding blood, using fiberoptic or rigid bronchoscopy, or other airway techniques.

Extinguish the bleeding: Bronchial artery embolization, especially using the superselective technique, has emerged as the most reliable way of locating and controlling bleeding. Immediate surgery should be reserved for life-threatening hemorrhage when the patient remains unstable, and for major vascular catastrophes (e.g., large convoluted AVMs, bronchovascular fistulas and PA rupture or penetrating lung trauma). Once control of hemorrhage has been achieved definitive management, depending on the etiology, can be instituted.

Watch for recurrences: immediate control may be achieved in 88-94% of occurrences, but recurrences are common, ranging from 10-28% after embolization in recent reports.

A key element of this approach is to attempt to lateralize the source of bleeding urgently, both to protect the opposite lung and to facilitate a more focused approach to embolization.

2. Emergency Management

For all patients: supplemental oxygen, 2 large-bore IV lines, type and cross, fluid resuscitation, or blood transfusion (even with un-cross-matched blood) as indicated, place with side of origin of bleeding (if known) dependent, obtain immediate labs as described below and correct coagulopathy if present. Give 15-30 mg IM codeine to reduce coughing, if no contraindications, and consider
small dose of anxiolytic, if ventilation not at risk, to alleviate anxiety (e.g., 0.25-0.5 mg lorazepam IV, or 2.5 mg IV morphine sulfate).

Exclude hematemesis (hx of abdominal pain, nausea, emesis, retching rather than coughing, dark, acidic blood), epistaxis (e.g., hx of blood from nostrils, underlying nasal pathology, recent/recurrent cocaine use, sensation of post-nasal drip, evaluate nasopharynx visually or with laryngoscope/nasopharyngoscopy, if needed) and distinguish from true hemoptysis (up to two thirds of patients will be able to localize bleeding to a site in the thorax with vague sensations of gurgling or fullness, bright red and frothy blood, alkaline pH, mixed with sputum, or rales on exam often unilaterally). Perform a quick assessment of surgical candidacy. Admit all patients with major bleeding to ICU.

If bleeding is severe enough to possibly require embolization, strongly consider institution of prophylaxis against contrast nephropathy with IV N-acetylcysteine and IV fluids, since multiple episodes of angiography may be required and renal function may be intrinsically compromised by hemodynamic instability.

Rarely, if bleeding is extremely brisk, intravenous terlipressin or vasopressin could be considered in patients without hypertension or ischemic heart disease to vasoconstrict bronchial arteries, but this carries systemic risks and limits the effectiveness of subsequent bronchial angiography and embolization due to the vasoconstriction.

Intravenous administration of tranexamic acid has been used to slow bleeding.

If a major bronchus is tamponaded with a bronchial blocker to contain bleeding, consider broad-spectrum antibiotics to reduce risk of pneumonia, a well-described complication.

For less severe bleeding: described as bleeding where the volume or frequency of bleeding is unlikely to result in compromised airway patency, severe hypoxia or cardiovascular collapse, most clinicians employ early (within 48 hours) fiberoptic bronchoscopy to locate the presumed segment of bleeding and identify any endoluminal lesion that is amenable to bronchoscopic intervention. Nevertheless, utility of bronchoscopy has been questioned: misses bleeding source in up to 20%, infrequently provides additional clinical information over CT scanning or changes the plan of management.

Contrast-enhanced CT scanning using multidetector modality can identify parenchymal abnormalities, abnormal vessels, AVMs, pseudoaneurysms, and aberrant vasculature likely to be the source of bleeding in most cases. If diffuse alveolar hemorrhage is suspected, bronchoscopy with lavage is instrumental in making the diagnosis, with increasingly bloody return after each aliquot of lavage, and demonstration of >30% hemosiderin-laded macrophages in the cellular sediment.

For massive hemoptysis: defined as bleeding where the volume and/or frequency of bleeding imminently impairs airway patency or oxygenation or provokes hemodynamic instability, immediate intubation is vital.

Intubation with a double-lumen endotracheal tube is not recommended: requires an experienced operator, tubes are often misplaced or displaced leading to aspiration and death, can be left in place only for a limited duration, can become occluded by blood clot due to narrow luminal caliber and preclude all fiberoptic bronchoscopy and bronchial manipulation due to the narrow lumen.

Intubation should be with a single-lumen ET tube > size 8 diameter; should preferably be performed with the patient in the lateral position with affected side down. Newer ET tubes are available with a built-in balloon catheter that can be used for bronchial tamponade.

After intubation is completed, a fiberoptic bronchoscope is passed through the ET tube and under visual guidance the ET tube is rotated to allow the balloon catheter to be advanced into either mainstem bronchus and inflated, successfully tamponading the involved lung. On the right side this results in right upper lobe collapse, worsening hypoxemia, and so should be avoided, using selective intubation of the left mainstem bronchus instead.

Intubation with a rigid bronchoscope is emphatically preferred in situations where bleeding is truly massive: provides efficient tracheobronchial toilet, better visualization of the airways, more effective tamponade and isolation of affected lung, and easier institution of coagulation therapies than does fiberoptic bronchoscopy. In addition, fiberoptic bronchoscopy can be combined with rigid bronchoscopy. However, this expertise is not widely or readily available.

Algorithm for massive hemoptysis

Intubate followed by rigid bronchoscopy if available; otherwise fiberoptic bronchoscopy with endobronchial manipulations as needed to tamponade bleeding (definitive therapy by bronchoscopy for reachable, visualized endobronchial lesions). When patient is sufficiently stabilized, proceed to CT scan, preferably multidetector CT scanning, followed by bronchial arterial embolization. If no bleeding source is visualized, evaluate non-bronchial systemic vessels.

Perform pulmonary angiography if initial bronchial embolization fails since this may be source of early recurrent bleeding in up to 93% in this circumstance. Bronchial arterial embolization can be safely reattempted. Reserve surgery for severely unstable patients despite bronchial tamponade therapies not permitting time for embolization, technical failure of embolization, recurrent hemoptysis despite repeat embolization attempts, or lesions requiring surgery for definitive management (e.g., vascular rupture, bronchovascular fistulas, tumor, large AVM, chest trauma, hydatid cyst or mycetoma).

Bronchial therapies for control of bleeding

Topical treatments

Iced saline lavage in 50-mL aliquots: shown to temporize or slow bleeding when used with vigorous airway toilet using a rigid bronchoscope

Instillation of topical epinephrine (1:20000), terlipressin and ornipressin have been used but carry the risk of systemic absorption and adverse cardiac effects, and the drugs may not be able to be delivered effectively because of vigorous bleeding.

Topical tranexamic acid has been applied topically with cessation of bleeding not otherwise controlled with the above treatments, as has topical application of fibrinogen-thrombin. The latter may be particularly effective as a short-term measure.

Tamponade measures

Fogarty balloon catheters placed into a mainstem bronchus to allow tamponade of bleeding and preserved ventilation of the contralateral lung. A new double-lumen Fogarty balloon tamponade catheter can be placed through the working channel of the fiberoptic bronchoscope and allows instillation of topical agents and suctioning. Even a PA balloon catheter has been used similarly and can be placed distally into segmental bronchi. Other mechanical methods of obtaining tamponade include placement of a silicone spigot, sealing of distal bleeding bronchi with transcatheter instillation of n-butyl cyanoacrylate, or bronchoscopy-guided forceps placement of oxidized cellular mesh directly into the orifice of the bleeding segmental bronchus.

Definitive therapy of endoluminal lesions reachable by bronchoscope includes use of Nd-YAG laser or argon plasma coagulation to control bleeding and is usually employed in patients with obstructing endobronchial malignancies, bleeding telangiectasias or catamenial hemorrhage.

3. Diagnosis

History and exam

Patients can report side of bleeding with 70-100% accuracy, physician exam accurate 43% and equivocal 25%.

History of active or prior TB, bronchiectasis, pneumonia, smoking history indicating possibility of lung cancer, prior epistaxis from Wegener’s, aphthous or genital ulceration in Behcet’s, prior pneumothorax in women of childbearing age suggesting lymphangioleiomyomatosis, multiple miscarriages in antiphospholipid syndrome, exposure to drugs or inhalational injury, cough and stridor or platypnea deoxyia syndrome indicating proximal airway tumor, perimenstrual hemoptysis, travel to areas endemic for lung parasites, etc.

Physical exam for characteristic stigmata: e.g., submucosal telangiectasias in HHT, nasal perforations or uveitis in Wegener’s granulomatosis or Behcet’s respectively, purpura on legs or buttocks in Henoch-Schonlein purpura, large tongue in amyloidosis, S3 or mid-diastolic murmur with opening snap, cyanosis and clubbing in congenital heart disease, leg swelling indicative of DVT and possible pulmonary embolism, etc.

Chest X-ray: Indicates side of bleeding in 33-82% and lesion of source in 35%, but may be normal. Negative x-ray suggests bronchiectasis as cause. Not uncommonly, x-ray is unreadable due to obscuration by blood.

CT scan: efficiency of localization is 70-90%, detects lesion from which bleeding is originating more efficiently than bronchoscopy (60-80 vs 2-8%). Multi-detector CT scans, if available, accurately detect abnormal bronchial vessels and distinguish pulmonary arterial origin of bleeding, obviating need for bronchial arteriograms. Not applicable for unstable patients, major bleeding requiring airway toilet or bronchial tamponade or if diffuse alveolar hemorrhage suspected.

Fiberoptic bronchoscopy: Utility in non-massive hemoptysis is questionable; often does not localize the lesion. In massive hemoptysis, useful where rigid bronchoscopy is not available, to localize source and attempt tamponade. Once that is achieved, CT remains superior in detecting bleeding source.

An anatomical list of the causes of hemoptysis
  • Larger airways: aspirated foreign body, broncholith, endobronchial tumor (carcinoid, sarcoma, adenoma, carcinoma primary or secondary), bronchial rupture (trauma), bronchovascular fistulas at level of aorta or PA (trauma, post-lung transplant anastomotic complications, mycotic, infectious pseudoaneurysms), bronchial telengiectasias (HHT), bronchial Dieulafoy lesions
  • Distal airways: Bronchiectasis (tuberculosis, CF, other), bronchitis, catamenial (endometriosis)
  • Parenchyma: Cavity (cancer, abscess, sarcoid, aspergilloma/ mycotic, vasculitic, parasitic [e.g., hydatid cyst]). Cysts: lymphangioleimyomatosis. Consolidation (necrotizing pneumonia, lung cancer or metastases, sarcoid, lymphoma, carcinoma, pulmonary infarction), Tumor, parasitic infections, heart failure, mitral stenosis, congenital heart disease.
  • Larger vessels: AVMs largely in patients with HHT, pulmonary artery rupture from PA catheter, aneurysmal rupture from mycotic, infectious, tubercular or vasculitic (e.g., Behcet’s, Hugh Stovin syndrome) aneurysms 5. Diffuse alveolar hemorrhage (vasculitides such as Goodpasture’s syndrome, Wegener’s, ANCA-associated vasculitis, Henoch-Schonlein purpura; connective tissue disorders such as SLE; coagulation disorders such as antiphospholipid syndrome, DIC; drugs (nitrofurantoin, cocaine, trimellitic anhydrides, solvent inhalation); other: idiopathic pulmonary hemosiderosis, amyloidosis, pulmonary capillary hemangiomatosis, pulmonary veno-occlusive disease
Lab tests

Immediate: CBC,type and cross, PT, PTT, platelet count, fibrinogen level/ DIC panel if relevant, ABG and chest x-ray, BUN/Cr as a baseline and together with UA if a pulmonary-renal syndrome is suspected, and serology for pulmonary or systemic vasculitis if diffuse alveolar hemorrhage is suspected, sputum Gram stain and culture, AFB stain and fungal stains and mycobacterial cultures. Baseline EKG and troponin.

After stabilization, and prior to angiography if the patient’s condition allows and time permits: CT chest with contrast, preferably using multidetector CT scanning. If bleeding is severe and patient is unstable, immediate fiberoptic bronchoscopy may help localize the segment of origin, which guides focused bronchial arterial embolization.


The lungs have a twin blood supply from bronchial arteries arising from the aorta, and the pulmonary arteries. The bronchial arteries supply the proximal airways and drain into the bronchial veins, while the distal airways to the level of terminal bronchioles are supplied by bronchial arterioles that form anastomoses with pulmonary arteries and drain into the pulmonary veins. These often hypertrophy during inflammatory conditions and form collaterals or elicit neovascularization that complicates attempts at embolization when they are the source of bleeding.

90% of hemoptysis originates from the bronchial arteries because these vessels are at systemic pressures and supply the majority of the airways.

Dieulafoy lesions are ectatic submucosal bronchial arterioles that have been described, most often in the proximal right bronchus, as a cause of massive hemorrhage, and may involve a malformation that causes a left-to-right shunt. AVMs in the Osler Weber Rendu syndrome are most often intrapulmonary.

Only 5% of hemoptysis is attributable to the pulmonary arteries, which have a pressure that is a fifth of systemic, but infectious and inflammatory pulmonary artery pseudoaneurysms are a common finding (11%) on angiographies conducted for hemoptysis. Similarly, Rasmussen’s aneurysm arises from pulmonary arterioles traversing tuberculous cavities, and they are often the source of bleeding in this condition.

The remaining 5% of bleeding arises from areas of lung pathology supplied by non-bronchial systemic arteries (e.g., subclavian, axillary, intercostal, internal mammary and phrenic).

In diffuse alveolar hemorrhage, majority of blood is usually accommodated within the lung such that hemoptysis is not profuse, and is often a later symptom, after anemia and hypoxia.

The source of bleeding, despite extensive investigation, remains undetermined in up to a fifth of cases.

Bronchial arteries arise from the aorta most often between vertebral levels T5-T6, but this can vary. The most common pattern (40%) is for a right bronchial artery to arise from an intercostobronchial trunk, and two independent left bronchial arteries arising from the aorta, but nine other patterns have been described, including origin of right and left bronchial arteries from a common trunk.

A full 16-30% of bronchial arteries arise from aberrant systemic arteries (e.g., brachiocephalic, subclavian, inferior phrenic, internal mammary, thyrocervical or costocervical trunks, celiac or left gastric or abdominal aorta), illustrating the complexity of performing successful bronchial embolization. In 5% the anterior medullary artery supplying the spinal cord can originate from the right intercostobronchial trunk. The risk of spinal cord infarction from embolization of bronchial artery is real but greatly reduced by awareness of this anatomical variation and use of the superselective technique to advance a microcatheter beyond the origin of the spinal artery.


Very common symptom (6.8% of office visits, 11% of hospitalizations) but massive hemoptysis accounts for only 5-15% of all cases of hemoptysis, although the prevalence in one series was as high as 39%. Tuberculosis, both active and remote, accounted for 78% of cases in the older literature and remains most common cause in endemic areas. Currently, chronic bronchitis-bronchiectasis most common, together with TB, necrotizing pneumonia and lung cancer account for the vast majority of cases.

Mortality from massive hemoptysis in recent series is lower (9-38%). Use of an approach with embolization first and then delayed surgery reduced hospital mortality, in comparison to an earlier historical cohort employing early surgery, from 15% to 0% in one series.

Risk factors for death: massive bleeding = 1000 mL/24 hrs, massive aspiration into contralateral lung, hemodynamic collapse, need for blood transfusions, need for single or independent lung ventilation, lung cancer as a source, recurrent bleeding after bronchial embolization, poor physical status to tolerate surgical intervention.

Risk factors for recurrence: residual bleeding in first week after embolization, need for blood transfusion after the initial bleed implying greater severity of hemorrhage, mycetoma.

Mortality rates from emergent surgical intervention range from 10% to 38%.

Morbidity from surgery deferred until after attempts at embolization in one recent report was 18%, with zero mortality. There were no deaths in another series in which surgical resections were performed only after embolization.

What’s the evidence?

Dweik, RA, Stoller, JK. “Role of bronchoscopy in massive hemoptysis”. Clin Chest Med. vol. 20. 1999. pp. 89-105. (Comprehensive review of management, epidemiology, outcomes and management of massive hemoptysis.)

Jean-Baptiste, E. “Clinical assessment and management of massive hemoptysis”. Crit Care Med. vol. 28. 2000. pp. 1642-7. (Excellent review of the management of massive hemoptysis.)

Shigemura, N, Wan, IY, Yu, SC, Wong, RH, Hsin, MK. “Multidisciplinary management of life-threatening massive hemoptysis: a 10-year experience”. Ann Thorac Surg. vol. 87. 2009. pp. 849-53. (Recent report emphasizing the reduction in mortality and morbidity from delaying emergent surgery until after bronchial embolization, and reduction of morbidity by utilizing rigid bronchoscopy for airway toilet intra- and pre-operatively. Compelling evidence for using a multidiscplinary approach.)

Sakr, L, Dutau, H. “Massive hemoptysis: an update on the role of bronchoscopy in diagnosis and management”. Respiration. vol. 80. 2010. pp. 38-58. (An exhaustive, up-to-date and excellent review of the bronchoscopic management of massive hemoptysis.)

Chun, JY, Morgan, R, Belli, AM. “Radiological management of hemoptysis: a comprehensive review of diagnostic imaging and bronchial arterial embolization”. Cardiovasc Intervent Radiol. vol. 33. 2010. pp. 240-50. (Comprehensive review of bronchial arterial embolization techniques as first-line therapy for management of massive hemoptysis.)