Management and prognosis of parapneumonic pleural effusion and empyema in adults

INTRODUCTION — A parapneumonic effusion is a pleural effusion that forms in the pleural space adjacent to a pneumonia. When microorganisms infect the pleural space, a complicated parapneumonic effusion or empyema may result. Prompt therapy of these entities can be lifesaving.

The treatment and prognosis of parapneumonic effusion and empyema are reviewed here. The clinical diagnosis of parapneumonic effusion and empyema in adults and management in children are discussed separately. (See "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults" and "Epidemiology, clinical presentation, and evaluation of parapneumonic effusion and empyema in children" and "Management and prognosis of parapneumonic effusion and empyema in children".)

DEFINITIONS — A parapneumonic effusion refers to the accumulation of fluid in the pleural space in the setting of an adjacent pneumonia.

●An uncomplicated or simple parapneumonic effusion refers to a free-flowing effusion that is sterile.

●A complicated parapneumonic effusion refers to an effusion that has been infected with bacteria or other microorganisms (eg, positive Gram stain or biochemical evidence of marked inflammation). (See "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults", section on 'Complicated parapneumonic effusion and empyema'.)

●An empyema refers to a collection of pus within the pleural space, which can develop when pyogenic bacteria invade the pleural space, from an adjacent pneumonia, direct inoculation (eg, from blunt trauma) or other source. Empyema that develops from an adjacent pneumonia is a subclass of a complicated parapneumonic effusion. While a complicated parapneumonic effusion and empyema represent a spectrum of infection within the pleural space, no pus is directly visualized in patients with a complicated parapneumonic effusion.

●A complex effusion refers to an effusion with internal loculations (septae).

●A uniloculated effusion is one where the effusion is without internal septae (it is not necessarily free-flowing).

GENERAL APPROACH TO MANAGEMENT — The management of parapneumonic effusions and empyema generally includes prompt antibiotic initiation and drainage of infected pleural fluid.

●For most patients with known or suspected parapneumonic effusions or empyema, we start empiric antibiotics immediately. Antibiotic selection varies based on the site of acquisition (ie, community versus hospital-acquired), severity of illness, local epidemiology, and patient risk factors for drug-resistant pathogens or infection with other specific organisms.

In general, empiric regimens should include antibiotics that target anaerobes and other likely pathogens (eg, streptococci if community-acquired; methicillin-resistant Staphylococcus aureus [MRSA] and Pseudomonas if hospital-acquired) when a complicated parapneumonic effusion or empyema is suspected. Patients with known or suspected uncomplicated effusions can generally be treated similarly to other patients with CAP. (See 'Antibiotic therapy' below.)

●Because of the high morbidity and mortality associated with acute pneumonia and infected pleural effusions, antibiotics should not be delayed pending diagnostic testing or drainage of the effusion. However, an exception includes selected stable patients with indolent illness onset who lack signs or symptoms of systemic infection, it is reasonable to drain the effusion and send for microbiologic testing before starting antibiotic treatment. The spectrum of pathogens that cause subacute and chronic pleural effusions and empyema differs from acute empyema (eg, includes mycobacteria, fungi). Deferring antibiotic therapy until microbiologic testing has been obtained may enhance diagnostic yield and allow for targeted therapy.

The approach to drainage depends on the type, size, and complexity of the effusion.

●For patients with small uncomplicated parapneumonic effusions (ie, sterile effusions), drainage is generally not necessary unless the effusion is sizeable enough to impair respiratory function. Close clinical and radiographic monitoring should be performed to ensure that the effusion is resolving. Larger effusions have increased risk of complications (algorithm 1). (See 'Uncomplicated parapneumonic effusion (antibiotics alone)' below.)

●For patients with complicated parapneumonic effusions (ie, with clinical or laboratory evidence of infection) or empyema, drainage should be performed as soon as possible for source control (algorithm 2). This is particularly true for empyema, which carries a worse prognosis. Loculated effusions, large free-flowing effusions (eg, ≥0.5 hemithorax), and effusions with a thickened pleural membrane should also be drained. When the collection is free-flowing, a single tube or catheter thoracostomy is the procedure of choice. When the collection is loculated, the approach to drainage is individualized depending on the complexity of the effusion and the patient's severity of illness. A common approach is to place a single tube or catheter in the largest locule and reassess the need for placement of additional drains and/or surgical intervention based on clinical and radiographic response. (See 'Complicated pleural effusion and empyema (antibiotics plus drainage)' below.)

Because a substantial proportion of patients require more than one procedure to achieve adequate evacuation of the effusion and lung reexpansion, the need for close clinical monitoring and serial imaging to assess treatment response cannot be overstated [1]. (See 'Assessment of response' below and 'Treatment failure' below.)

Our approach is generally similar to that outlined by the American Association for Thoracic Surgery (AATS), the European Association for Cardio-Thoracic Surgery (EACTS), the American College of Chest Physicians (ACCP), the British Thoracic Society (BTS), and the European respiratory Society/European Society of Thoracic Surgeons  [2-6]. However, the original categories outlined by the ACCP have largely fallen out of favor.

ANTIBIOTIC THERAPY — In general, antibiotic therapy mirrors that selected for the underlying pneumonia. However, attention should also be paid to the appropriate coverage of anaerobic bacteria and to choosing antibiotics that have good penetration into the pleural space.

Empiric therapy (agent choice) — For most patients, empiric antibiotic therapy should be started as soon as the diagnosis of a parapneumonic effusion or empyema is known or suspected. While drainage of infected fluid within the pleural space is critical to care, antibiotic initiation should not be delayed while awaiting diagnostic procedures (eg, thoracentesis) or drainage. Exceptions can be made for selected stable patients with long-standing effusions since the pathogens that cause subacute and chronic empyema differ from those associated with acute pneumonia (eg, mycobacteria and fungi); in such situations, deferring antibiotic therapy until microbiologic testing has been obtained may enhance diagnostic yield and allow for targeted therapy.

Generally, empiric antibiotic regimens should include an antibiotic that targets anaerobic bacteria, which are common causes of complicated parapneumonic effusions and empyema. Additional antibiotics should be selected based on the site of acquisition (eg, community- versus hospital-acquired), mode of acquisition (eg, aspiration, trauma), and local epidemiology. While data suggest that influenza may precede pneumonias that develop pleural space infections, no data suggest that empiric treatment of influenza is indicated [7].

Nearly all antibiotics adequately penetrate the pleural space. Aminoglycosides (eg, gentamicin, amikacin, tobramycin) are exceptions. Because their pleural penetration is poor and because they may be inactivated in acidic environments (eg, empyemas), we generally avoid them when alternatives are available [8].

Initial antibiotic therapy should be given intravenously. Transition to oral therapy can be considered once the patient has demonstrated clear clinical improvement and adequate drainage has been achieved. There is no role for routine use of intrapleural antibiotics.

Community-acquired — For most community-acquired complicated parapneumonic effusions or empyema, we select an empiric IV antibiotic regimen that targets Streptococcus pneumoniae and the pathogens that colonize the oropharynx, including microaerophilic streptococci (eg, S. anginosus, S. intermedius) and anaerobic bacteria (table 1). Reasonable options include:

●A third-generation cephalosporin (eg, ceftriaxone or cefotaxime) plus metronidazole

●A beta-lactam/beta-lactamase inhibitor combination (eg, ampicillin-sulbactam)

For patients with penicillin hypersensitivity who cannot tolerate cephalosporins (algorithm 3), alternate options include monotherapy with a carbapenem (eg, imipenem, meropenem), combination therapy with a respiratory fluoroquinolone (eg, levofloxacin, moxifloxacin) plus metronidazole or a monobactam (eg, aztreonam) plus metronidazole. Although clindamycin has been used historically for the treatment of anaerobic lung infections, resistance rates to clindamycin among anaerobes now consistently exceed 20 percent across treatment settings. For this reason, we no longer routinely use clindamycin for empiric treatment of anaerobic infections. (See "Anaerobic bacterial infections", section on 'Antimicrobial resistance'.)

Modifications to these regimens may be needed for severely-ill patients or for those with risk factors for specific pathogens. As examples, we may expand coverage to include methicillin-resistant Staphylococcus aureus (MRSA) in patients with antecedent influenza infection or other MRSA risk factors (table 2). For patients with concurrent necrotizing community-acquired pneumonia (CAP), we may include coverage for both MRSA and Pseudomonas.

Local epidemiology should also be taken into account when selecting empiric antibiotics because the prevalence of pathogens that cause parapneumonic effusions vary with geography. Prominent examples include Burkholderia pseudomallei (cause of melioidosis) in Southeast Asia and tuberculosis in developing regions of the world. (See "Melioidosis: Treatment and prevention" and "Tuberculous pleural effusion".)

In general, we do not include an agent that targets atypical pathogens (eg, Legionella, Chlamydia, Mycoplasma spp) in our empiric treatment regimens, as these pathogens rarely cause complicated parapneumonic effusions and empyema. (See "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults", section on 'Microbiology'.)

Generally, patients with uncomplicated parapneumonic effusions can be treated similarly to other patients with CAP. (See "Overview of community-acquired pneumonia in adults", section on 'Treatment'.)

Hospital-acquired — For most hospital-acquired infections (eg, empyema secondary to health care-associated pneumonia or postprocedural empyema), we select an empiric IV antibiotic regimen that targets MRSA, gram-negative bacteria (including Pseudomonas spp), and anaerobic bacteria (table 1). For example, combining vancomycin with metronidazole and an antipseudomonal cephalosporin (eg, cefepime, ceftazidime) is appropriate. Combining vancomycin with an anti-beta-lactam/beta-lactamase inhibitor (eg, piperacillin-tazobactam, ticarcillin-clavulanate [limited supply]) is an alternative. However, there is growing concern that the combination of vancomycin plus piperacillin-tazobactam is nephrotoxic [9]. Thus, some clinicians use linezolid in place of vancomycin when piperacillin-tazobactam is used. For those who are penicillin-allergic, we suggest combining vancomycin with metronidazole and an antipseudomonal fluoroquinolone (eg, ciprofloxacin); alternatively, combining vancomycin with an antipseudomonal carbapenem (eg, imipenem or meropenem) is appropriate.

Others — The treatment of less common causes of parapneumonic effusions and empyema, such as Candida species, other fungi, melioidosis, and tuberculous empyema is discussed separately. (See "Candida infections of the abdomen and thorax", section on 'Empyema' and "Tuberculous pleural effusion", section on 'Management' and "Cryptococcus neoformans infection outside the central nervous system" and "Cryptococcus neoformans infection outside the central nervous system", section on 'Pulmonary infection in immunocompromised adults' and "Melioidosis: Treatment and prevention".)

Directed therapy — Definitive therapy should be based on culture results and clinical suspicion for a monomicrobial or a polymicrobial infection.

●When suspicion for a monomicrobial infection is high (ie, isolation of S. pneumoniae or S. aureus as sole pathogens from pleural fluid), it is reasonable to direct therapy at the isolated pathogen.

●In most other circumstances (eg, aspiration pneumonia, negative cultures, or isolation of a constituent of the oral/gastrointestinal [GI] flora such as S. milleri), we consider the infection to be polymicrobial and inclusive of multiple anaerobic bacteria. In these circumstances, we select a regimen based on the likely source of infection (eg, health care-associated pneumonia, community acquired pneumonia, or aspiration) that also targets the isolated pathogen as well as any other anaerobic bacteria. Because anaerobic bacteria can be difficult to culture and are common causes of parapneumonic effusions and empyema, most experts include an antibiotic that targets anaerobes for the duration of therapy regardless of culture results.

Duration of therapy — The optimal duration of therapy is not known. We generally individualize the duration of therapy based upon the type of effusion, the adequacy of drainage, clinical and radiographic response to treatment, and the patient's immune status.

In general, for self-resolving uncomplicated bacterial parapneumonic effusions, therapy may last one to two weeks, while therapy for complicated parapneumonic effusions and empyema are often longer (eg, two to three weeks for a complicated parapneumonic effusion and four to six weeks for empyema). While we take radiographic response into account when determining the duration of therapy, complete radiographic resolution may take many weeks or months and residual pleural thickening can persist for longer periods. Thus, treating with the goal of complete radiographic resolution is not necessary.

The initial IV antibiotic regimen can be switched to an oral regimen with a similar treatment spectrum when clinical response is clear (eg, patient is afebrile, hemodynamically stable, clinically improving), no further drainage procedures are needed, and the patient is able to tolerate oral medications.

The duration of antifungal and antituberculous therapies for empyema is discussed separately. (See "Candida infections of the abdomen and thorax", section on 'Treatment' and "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults without HIV infection" and "Treatment of pulmonary tuberculosis in adults with HIV infection: Follow-up after initiation of therapy", section on 'Duration of therapy'.)

APPROACH TO DRAINAGE

Uncomplicated parapneumonic effusion (antibiotics alone) — Uncomplicated parapneumonic effusions are small to moderate-sized (ie, less than half the hemithorax) free-flowing effusions with no evidence of infection by culture or chemistry [2] that generally resolve with antibiotics alone and generally do not need drainage. In such cases, the diagnosis and therapy with antibiotics alone are empiric (algorithm 1).

In some cases, thoracentesis may be performed. For example, if the effusion is sizeable enough to impair respiratory function (eg, typically in patients with underlying lung disease), drainage can be performed for symptomatic relief. Other indications may include patients with a severe clinical presentation, or patients in whom the pleural space is the suspected source of infection. If after thoracentesis, suspicion remains for infection in the pleural space despite a negative Gram stain, culture, or pleural fluid chemistries (eg, patient with septic shock), we generally proceed with drainage and treat the patient as if they have a complicated (ie, infected) parapneumonic effusion. (See 'Complicated pleural effusion and empyema (antibiotics plus drainage)' below.)

All patients with uncomplicated parapneumonic effusion should be followed clinically and with serial chest radiographs or ultrasound examinations to assess for improvement or deterioration. The optimal frequency of radiographic follow-up is unknown but it is appropriate that the first follow-up imaging be obtained within 48 hours if thoracentesis was not performed. If thoracentesis was performed and confirms an uncomplicated parapneumonic effusion, serial radiographs can be repeated within one week of the diagnosis and followed every one to two weeks until resolution since progression to empyema while appropriate antibiotics are being administered is rare. Details regarding the diagnosis of an uncomplicated effusion are provided separately. (See 'Definitions' above and "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults", section on 'Uncomplicated parapneumonic effusion'.)

Should patients fail to improve, the effusion enlarges, or new fever develops, repeat imaging with chest computed tomography should be performed to evaluate for the development of a complicated parapneumonic effusion that may need to undergo sampling and drainage. (See "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults", section on 'Diagnostic evaluation'.)

Complicated pleural effusion and empyema (antibiotics plus drainage) — In addition to appropriate antibiotic therapy, PROMPT drainage is indicated in patients when there is clinical concern for or evidence of infection in the pleural space, based upon the following features (algorithm 2):

●Empyema (ie, overtly purulent pleural fluid)

●Positive pleural fluid Gram stain or culture

●Loculated pleural effusion

●Large free-flowing effusions (ie, ≥0.5 hemithorax)

●Effusions associated with thickened parietal pleura

●Sepsis from a pleural source

This approach is based upon the rationale that without drainage (ie, source control), patients have poor outcomes including an increased requirement for more than one procedure, eventual need for surgery, and longer hospitalization. This is particularly important for empyema, which carries the poorest prognosis and highest mortality.

A pleural fluid pH of <7.2 is also an indicator of infection in the pleural space. However, other pleural diseases can have a low pleural fluid pH (eg, malignant effusions, rheumatoid and lupus pleurisy, urinothorax, and saline from a misplaced central venous catheter) (see "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults", section on 'Differential diagnosis'). Therefore, the decision to drain fluid from the pleural space based on a low pleural fluid pH alone should be made after pleural fluid analysis is complete.

The initial procedure of choice is a single tube or catheter thoracostomy. Importantly, this recommendation applies to those in whom residual effusion remains following diagnostic thoracentesis. However, when an effusion is loculated, choosing to drain the largest locule (usually guided by ultrasound or chest computed tomography [CT]) is appropriate; in such situations, consideration should be given to the prompt insertion of a second or third drain during follow-up. Early thoracic surgical consultation is appropriate because some of these patients will require thoracoscopic or open surgery [3-5,10]. Diagnosis of an uncomplicated effusion is discussed separately. (See 'Definitions' above and "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults", section on 'Uncomplicated parapneumonic effusion' and 'Intrapleural tPA +/- additional drainage' below.)

Initial drainage (tube or catheter thoracostomy) — Chest tube or catheter thoracostomy drainage is the least invasive option for drainage of infected pleural fluid in patients with a complicated parapneumonic effusion or empyema. It is best suited for patients with free-flowing or uniloculated effusions (ie, effusion without internal septae), but is also frequently used to drain complex effusions (ie, effusions with internal septations or locules). Details of tube or catheter thoracostomy management that are specific to patients with parapneumonic effusion and empyema are discussed in this section. The technique of insertion, general management, and complications of tube thoracostomy are discussed separately. (See "Thoracostomy tubes and catheters: Indications and tube selection in adults and children".)

●Image guidance – When draining infected pleural fluid, thoracostomy tubes are typically placed using either ultrasound or CT guidance. However, many experts place chest tubes blindly at the bedside, especially when the effusion is large or free-flowing.

●Size – There is no consensus on the size of chest tube or catheter for drainage and guidelines vary on their recommendations [2-5,11]. In general, we prefer small-bores tubes (10 to 14 French [Fr]) based upon data that suggest similar efficacy and less pain when compared with large-bore thoracostomy tubes [12]. However, in practice the choice may be dependent upon factors including physician and patient preference, institutional policy, and available expertise. Some experts prefer larger bore tubes in patients with effusions that have multiple locules since larger tubes may penetrate locules more readily than smaller tubes.

Traditionally, larger bore tubes (>28 Fr) were preferred for drainage of more viscous empyema fluid and smaller bore tubes were reserved for less viscous fluid. However, in a prospective, nonrandomized study of 454 patients who underwent chest tube drainage for empyema as part of the Multi-center Intrapleural Streptokinase Trial (MIST1), no significant difference was found in mortality or need for thoracic surgery between large (15 to 20 Fr), medium (10 to 14 Fr), or small (<10 Fr) bore tubes [13,14]. Pain was markedly less with the smaller sized tubes. However, data suggest that smaller bore tubes (up to 14 Fr) may be more prone to blockage with viscus empyema fluid as well as with blood or fibrinous debris, with one study reporting that drain occlusion occurs more commonly in those with empyema (11 to 30 percent) [15]. Given the propensity for small-bore catheters to become blocked, periodic flushing (eg, 30 mL of sterile saline every six hours via a three-way valve) may help to maintain tube patency [16]. Large-bore chest tubes do not need to be flushed (unless blockage is suspected) and often do not have a three-way tap for frequent flushing.

●Suction – The application of suction is typical to assure maximal and consistent pleural fluid removal since pleural fluid output is the major determinant that suggests that any tube can be removed. However, suction is not necessary unless the pleural space fails to drain or an air leak is present. Once an air leak is excluded, the chest tube or catheter can be placed to water seal. In rare cases, some catheters empty directly into a drainage bag without the option of suction, similar to percutaneous catheters commonly used for abdominal abscess drainage.

●Efficacy – Although some patients with complicated parapneumonic effusions may improve with antibiotics alone, the response is variable and drainage is not always successful. No controlled studies are available to guide selection of patients for drainage. All of the features that comprise a complicated effusion and empyema (loculations, large size, pleural fluid acidosis) are associated with an increased risk of progression and poor outcomes including the requirement for more than one procedure, eventual need for surgery, and longer hospitalization [17-20]. One meta-analysis of seven observational studies reported that pleural pH <7.2 was the most useful predictor of a complicated clinical course [21]. If pleural pH is not measured, a pleural fluid glucose value <40 mg/dL (2.2 mmol/L) and/or pleural fluid lactate dehydrogenase (LDH) value >1000 international units/L, or significant loculations also appear predictive of the need for tube thoracostomy. Thus, most experts agree that drainage is indicated in this population. Thoracic empyema (ie, pus in the pleural space) invariably requires drainage (akin to draining a pyogenic abscess for source control) because among those with an infected pleural space, patients with empyema have the highest mortality.

Assessment of response — Immediately following drainage, patients typically undergo chest radiographic imaging for a rudimentary assessment of tube or catheter placement and response. Patients are then followed clinically (signs and symptoms, laboratory values, and drainage volume) for the next 24 to 48 hours. Chest CT imaging is typically performed within that time frame since CT provides accurate details regarding chest tube position and adequate drainage of the effusion or empyema [22], thereby helping to inform the clinician regarding the next-steps in decision-making.

●If imaging shows that the chest tube is not in a good position, then it can be manipulated or replaced (typically under image guidance), following which drainage volume should be monitored and a repeat chest CT obtained.

●If the drainage tube is in good position but the drainage is inadequate and/or the lung has not reexpanded adequately, then several options need to be explored, the details of which are provided below. (See 'Treatment failure' below.)

●If imaging reveals marked improvement, removing the drainage tube should be considered, the details of which are discussed below. (See 'Discontinuing drainage' below.)

Bronchoscopy is rarely indicated unless there is suspicion for endobronchial obstruction or a bronchopleural fistula.

Discontinuing drainage — Chest tubes can generally be removed when drainage volume falls below 50 to 100 mL/day (for two to three days assuming there is no blockage), pleural imaging shows reasonable size reduction of the effusion, and the patient has no or resolving signs of clinical infection. Complete resolution of pleural thickening radiologically is not required as this may require months. The patient is typically discharged on antibiotics and evaluated as an outpatient with follow-up imaging (typically chest CT) in about two weeks. A small fraction of patients may be discharged with a temporary catheter in place for clearance of residual fluid, provided adequate follow-up in outpatient clinic is assured. Antibiotics should be discontinued once clinical and radiologic improvement is observed and it is assured that the effusion has not worsened or reaccumulated. Infectious disease experts are helpful in determining the point at which antibiotics can be stopped. (See 'Duration of therapy' above.)

Treatment failure — Failure to improve after antibiotics and tube thoracostomy drainage (eg, the effusion persists or worsens, fever persists or new fever develops, persistent or worsening leukocytosis) may indicate that antibiotic coverage and/or that drainage is inadequate. While some patients in this category require surgical intervention (see 'Video-assisted thoracic surgery (VATS)' below), many patients (50 to 80 percent) may respond to nonsurgical alternatives such as antibiotic adjustment, additional drainage procedures, and/or intrapleural tissue plasminogen activator (tPA) with /deoxyribonuclease (DNase). This approach is based upon the high success rate of tPA/DNase treatment which decreases the likelihood of intervention and shortens the duration of hospitalization. (See 'Assess adequate antibiotic coverage and drainage' below.)

Most experts consider sizeable residual loculations and poor lung reexpansion as evidence of treatment failure; others do not use these parameters as indicators of treatment failure but rather rely on clinical assessment only. The former approach is more aggressive and focused on a more rapid resolution of the complicated parapneumonic effusion, whereas the latter approach is based upon the rationale that many parapneumonic effusions should resolve over time with antibiotics alone. The risk of being nonaggressive is the possibility of residual trapped lung.

Assess adequate antibiotic coverage and drainage

●Assessing antibiotic coverage – For patients with evidence of ongoing or worsening infection (eg, new fever, increasing white blood cell count, worsening effusion, development of sepsis), options include reculturing pleural fluid directly from the pleural space (not from the tube or catheter) or undrained locule and adjustment of antibiotic coverage ensuring adequate anaerobic coverage or coverage of resistant organisms. Consideration should also be given to the presence of underlying necrotic lung as a focus of infection.

●Assessing adequate drainage – For patients in whom treatment failure is considered due to inadequate drainage, further drainage of the infected pleural space should be performed. Several drainage options exist, which are discussed in the section below.

Drainage options

Choosing among drainage options — Choosing among the drainage options is often provider-specific and recommendations vary among clinicians, institutions, and guideline committees. Factors that influence the decision include the clinician's specialty, expertise availability, patient values, patient prognosis from comorbidities, and candidacy for select surgical procedures as well as the number and size of locules and degree of pleural thickening on chest CT.

Some general rules can provide guidance:

●In most cases of treatment failure, when imaging reveals that the tube or catheter is in a good position but the effusion is only partially improved (ie, a significant amount of fluid remains) or is complex (ie, loculated) and poorly drained, most experts administer intrapleural tPA/DNase and give consideration to placing a second (or third) chest tube or catheter before resorting to video-assisted thoracic surgery (VATS). The rationale for this approach is that this strategy reduces the need for surgery and shortens the duration of hospitalization. A repeat chest CT should be performed 24 hours following the completion of the chosen intervention(s) and early thoracic surgical consultation is also advisable. If tPA/DNase is not an option, saline washes may be an alternative.  (See 'Intrapleural tPA +/- additional drainage' below.)

●Nonsurgical treatments may be the only option in those who are not surgical candidates; this is supported by one retrospective study that reported nonsurgical interventions occurred more frequently in those with medical comorbidities (eg, congestive heart failure, chronic lung disease, peripheral vascular disease, metastatic cancer, diabetes with complications, coagulopathy anemia of chronic disease, sepsis) [1]. (See 'Intrapleural tPA +/- additional drainage' below.)

●Patients with empyema from a bronchopleural fistula should preferably undergo surgical or endoscopic repair procedures after an adequate course of antibiotics have been administered. (See 'Postsurgical empyema' below and "Bronchopleural fistula in adults".)

●In patients who have clear evidence of significant organization or an established fibrothorax (eg, no pleural fluid, excessively thickened pleura on chest CT with or without calcification), or evidence of trapped lung (air replaces fluid after drainage), many experts proceed directly with VATS for decortication rather than attempt a trial of tPA/DNase. In such cases, surgery is not urgently needed and preoperative assessment with pulmonary function testing can proceed on an elective basis. (See 'Video-assisted thoracic surgery (VATS)' below and "Diagnosis and management of pleural causes of nonexpandable lung".)

Intrapleural tPA +/- additional drainage — Nonsurgical approaches involve the intrapleural instillation of fibrinolytics (typically tPA) together with DNase and/or the placement of additional chest tubes. Most experts attempt tPA/DNase first but no data exist to support choosing one over the other. The increasing use of fibrinolytics was reflected in a retrospective analysis of over 4000 patients undergoing an intervention for empyema, which reported that over half of patients underwent nonsurgical therapies (including antibiotics, thoracostomy tubes, and fibrinolytics); a quarter of these patients needed VATS and one-fifth needed open surgery, suggesting a 55 percent response rate to nonsurgical therapies [1].

●Intrapleural tissue plasminogen activator with DNase — The indications, dosing, follow-up, and efficacy of tPA/DNase are discussed in this section. It is important to administer both agents in combination rather than either agent alone.

•Indications – Intrapleural tPA/DNase is frequently administered to patients with complicated parapneumonic effusion or empyema who fail antibiotic therapy and initial drainage. It is also a suitable option in patients who are not candidates for or do not want surgery.

Predicting a response to this strategy is difficult. One study suggested that pleural thickening and the presence of an abscess/necrotizing pneumonia may help to predict failure [23]. In theory, this strategy is probably best suited to patients with multiloculated effusions in whom septae are suspected to be early and amenable to breakdown. In contrast, it is less effective for those with significant organization (ie, stage 3 disease (table 3)) (see "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults", section on 'Stage 3 (chronic organization)'). While a thickened pleural space may indicate significant organization, in reality, it can be difficult to determine how organized a parapneumonic effusion or empyema is by chest CT scan, or to determine whether locules are in communication with another; thus, many experts administer tPA/DNase empirically and decide whether surgery is needed depending on the response. Because there is often a diagnostic delay in recognition of a parapneumonic effusion, there are no data that define a time after which fibrinolytic/mucolytic therapy will not work.

•Dosing – Although several agents and several doses of fibrinolytics have been used (urokinase, streptokinase, tPA), most institutions use a regimen that is similar to that described in the largest randomized trial that showed a reduction in the need for surgery when tPA was combined with DNase [24]. Both tPA (10 mg) and DNase (5 mg) are administered via the chest tube or catheter twice daily for three days [24]. In the trial, tPA and DNase were administered separately and tubes were clamped for one hour after each agent had been given, although newer data suggest that the simultaneous administration of both agents may be as efficacious [25]. Retrospective series describe that regimens with lower doses of tPA may be as effective (eg, 5 mg tPA two times per day and 5 mg DNase two times per day, for 3 day)  and may be considered in patients at higher bleeding risk [26].

Once a course of tPA/DNase has been administered, it is not typically repeated, since this is thought to increase the risk of pleural bleeding. However, consideration may be given to administering a second course if no additional options are available and a response was noted previously. Options in refractory cases are discussed below. (See 'Refractory cases' below.)

•Follow-up – Follow-up involves assessing clinical findings, chest tube/catheter drainage volume, and radiographic imaging with chest CT. However, drainage volume can be a misleading measure of response because fibrinolytics can increase fluid production de novo, without decreasing empyema size.

-If the effusion improves dramatically or resolves and the drainage is minimal, then consideration should be given to removing the drainage tube. (See 'Discontinuing drainage' above.)

-If the patient experiences minimal or no response, then VATS is typically indicated. (See 'Video-assisted thoracic surgery (VATS)' below.)

-If the patient experiences a partial response, the decision regarding how to proceed is individualized. In our experience, much of the decision is dependent upon the presence of comorbidities, patient preferences, and the degree of clinical response to treatment. Those who demonstrate a significant response may not need to proceed to VATS, while those with a marginal response should probably be evaluated for VATS.

•Adverse effects – Side effects of intrapleural administration of fibrinolytic agents include chest pain, fever, allergic reactions (more frequent with streptokinase), and pleural hemorrhage [13,27-30].

Pleural hemorrhage is the most worrisome side effect, since blood loss can be significant. In addition, blood in the pleural space perpetuates the lack of visceral and parietal pleural apposition and increases the risk for lung entrapment; it is, therefore, frequently an indication for evacuation by VATS. The incidence ranges from one to seven percent. The risk for pleural hemorrhage is highest in those who are already at risk of bleeding (eg, renal failure, thrombocytopenia, anticoagulation) [29]. Thus, one should be cautious administering fibrinolytics in those with a known coagulopathy. Hemothorax will be detected relatively quickly by seeing blood in the thoracostomy tube output. Any coagulopathy should be reversed, if feasible, and thoracic surgery should be consulted for a VATS evaluation.

Fibrinolytics are quickly neutralized (usually within an hour) by plasminogen activator inhibitors that are increased during pleural infection [29,31,32]. Consequently, they rarely cause systemic bleeding, although a few case reports have been described [33,34].

•Efficacy – The administration of intrapleural fibrinolytic agents is based upon the rationale that they break down fibrinous adhesions that are part of the organization process and are responsible for locule (septae) formation. DNase, by breaking down DNA (released from tissue and bacteria), reduces the viscosity of pleural fluid. In theory, these agents, when administered together, should improve the efficacy of pleural drainage by thoracostomy tube/catheters. Data show that when administered together [24,25,35,36], they are more efficacious than when fibrinolytics are administered alone [37-43]. These agents typically decrease the need for surgery (by approximately 30 to 80 percent) but a mortality benefit has not been proven.

-Fibrinolytics plus DNase – One trial randomly assigned 210 patients with empyema to one of four intrapleural treatments: 10 mg intrapleural tPA twice daily for three days, 5 mg intrapleural DNase twice daily for three days, a combination of tPA and DNase twice daily for three days, or double placebo [24]. Combination tPA-DNase therapy resulted in a greater decrease in radiographic pleural opacity (-7.9 percent change in pleural opacity; 95% CI -13.4 to -2.4), an 83 percent reduction in the rate of surgical referral (odds ratio 0.17; 95% CI 0.03-0.87), and a shorter hospital stay (-6.7 days; 95% CI -12.0 to -1.9) compared with placebo. Neither of the individual agents performed better than placebo. Two episodes of pleural hemorrhage and one episode of hemoptysis were seen with combined tPA-DNase therapy, although the rate of adverse effects was not different across study groups. In preplanned subgroup analyses, there were no differences in the outcomes between those with purulent versus nonpurulent effusions, large- versus small-bore drainage tubes, or patients with evidence of loculations versus no loculations. On the basis of this study, most institutions use this protocol when administering tPA/DNase to patients with complicated parapneumonic effusion or empyema.

A Cochrane meta-analysis of 10 randomized trials found a similar reduction in surgical intervention with the use of tPA-DNase compared with placebo (odds ratio [OR] 0.37; 95% CI 0.21-0.68) [44]. There were fewer treatment failures (OR 0.16, 95% CI 0.05 - 0.58) but no clear mortality benefit was identified (OR 1.16, 95% CI 0.71-1.91). However, most of the trials were at high risk of bias, and the efficacy was diminished when only trials at low or unclear risk of bias were included in the analysis. Thus, per 1000 people, tPA-DNase may lead to 19 more deaths (36 fewer to 59 more), 115 fewer surgical interventions (150 to 55 fewer), and 214 fewer treatment failures (252 to 93 fewer).

-Fibrinolytics alone – While early retrospective studies suggested benefit [13,37-39,41,42], newer data suggest mixed or no benefit when fibrinolytic agents are administered alone [13,43]. One randomized, double-blinded trial of 454 patients parapneumonic effusion or empyema showed no benefit from streptokinase in any category, including mortality, referral for surgery, imaging outcome, or hospital stay [13]. A subsequent meta-analysis of seven trials that included 761 patients with parapneumonic effusion or empyema reported that compared with conservative therapy alone (tube thoracostomy), intrapleural fibrinolytics did not result in a mortality benefit (28 versus 33 percent; relative risk [RR] 1.08; 95% CI 0.69-1.68) but did reduce the need for surgical intervention (RR 0.63; 95% CI 0.46-0.85) [45]. However, not all included studies were blinded. Another 2012 meta-analysis showed similar results [46].

●Additional drainage tubes — In some patients in whom drainage is inadequate with a single chest tube or catheter, particularly those with complex (ie, multiloculated) effusions and those who fail fibrinolytic therapy, the placement of additional chest tubes is appropriate. In the majority of cases, a second (and occasionally third) catheter is placed under image-guidance, typically chest CT. Occasionally, replacement of a small-bore catheter with a larger chest tube is performed by some experts, but the efficacy of this approach is unproven. Placement of additional drainage tubes is more likely to succeed if there is a small number of locules (eg, two to three, especially if large) or if it is known or suspected that one particular locule is infected. However, it is unlikely to work if patients have multiple small locules which are difficult to access and the location of infected locule(s) is along the mediastinal pleura or interlobar fissures. Lung abscesses should not be drained since these usually drain by the bronchial route with adequate antibiotics. Interventional radiologists should take care to not place drainage catheters through the lung parenchyma. (See "Lung abscess in adults".)

Video-assisted thoracic surgery (VATS) — VATS is often indicated in symptomatic patients with parapneumonic effusion or empyema that fails to resolve with antibiotics, tube thoracostomy, and a course of tPA/DNase [10,47,48]. (See 'Treatment failure' above.)

VATS is preferred over open thoracotomy since outcomes are similar and morbidity and hospital length of stay is lower [49-51]. While some surgeons prefer to proceed directly with open thoracotomy in some cases (eg, patients with significant adhesions, greater visceral pleural thickness, or larger empyema cavity size), others prefer to start with VATS and convert intraoperatively to open thoracotomy [52-54]; for example, some patients in whom stage 2 disease is suspected (fibropurulent stage) who turn out to have components of stage 3 (chronic organization) may need an open procedure for complete decortication; conversion is also appropriate in those with intolerance of single lung ventilation, uncontrollable bleeding, or needing access to structures not amenable to VATS repair. Conversion rates of up to 44 percent have been reported (on average, conversion rates are approximately 10 percent) [49,50,52-55]. Conversion to thoracotomy was more common in patients with delayed referral (>2 weeks) for VATS and those who had gram-negative bacteria causing empyema. In some decortication cases, underlying necrotic lung is discovered, prompting parenchymal resection.

Data suggest that the efficacy of VATS for treating patients with parapneumonic effusion and empyema is mixed [52-61]. While several small series suggest that thoracoscopy is superior to intrapleural fibrinolysis [58-62], a meta-analysis of seven randomized trials reported that when compared with tube thoracostomy (with or without fibrinolytics), VATS did not result in a mortality benefit but did result in a reduced length of hospital stay [57]. Most studies report successful resolution of the effusion in response to VATS.

Postoperative pain is managed in a similar fashion to other thoracic surgeries and although epidural catheterization is generally avoided due to the risk of infecting the epidural space there are no data to support this potential adverse effect. The technique, postoperative care, and complications of VATS are discussed separately. (See "Overview of minimally invasive thoracic surgery".)

REFRACTORY CASES — In rare patients who fail antibiotics, tube thoracostomy, and fibrinolytic/mucolytic therapy, who are not candidates for VATS, options are limited but may include persisting with conservative care (antibiotics and multiple drains) and in some cases, performing an open-window thoracostomy.

An open thoracostomy involves a vertical incision through the chest wall with rib resection to permit open drainage at the inferior border of the empyema cavity. The empyema can drain directly through the chest wall if a skin flap is pulled into the opening (an Eloesser flap) or can drain through a chest tube that is gradually advanced outward as the tract closes. This process takes approximately 60 to 90 days. Although this procedure is less invasive than decortication, the risks of general anesthesia and the morbidity of prolonged chest tube drainage remain. Some experts use multiple daily dressing changes or a wound vacuum-assisted closure (VAC) device instead of a chest tube to facilitate drainage of the empyema, although a wound VAC may worsen a bronchopleural fistula, if present, and should be avoided under those circumstances.

SPECIAL POPULATIONS

Postsurgical empyema — Empyema can complicate resective thoracic surgery (ie, lobectomy or pneumonectomy) and often occurs in association with a bronchopleural fistula. These infections once relegated the patient to a life of continued open thoracostomy; however, other options of closure are now available to the experienced thoracic surgeon [4,5,63]. These include combinations of pleural space sterilization with antibiotic irrigation, space-filling muscle pedicles, thoracoplasty with rib resection, and surgical or endoscopic closure of a bronchopleural fistula and/or open thoracostomy incisions. Further details are provided separately. (See "Sequelae and complications of pneumonectomy", section on 'Pleural space complications' and "Bronchopleural fistula in adults".)

Infected malignant effusion with indwelling catheter — Patients with an indwelling tunneled catheter for drainage of a malignant pleural effusion (MPE) can develop pleural infection that can progress to empyema. These infections often are due to staphylococcal species and typically resolve on a prolonged course of directed antibiotic therapy while leaving the catheter in place [64].

COMPLICATIONS — Long-term complications of infection in the pleural space include residual pleural thickening on chest computed tomography (CT) and, when untreated, fibrothorax and pleural calcification. A thickened pleura (also known as a "pleural peel") generally begins to regress once the pleural infection is adequately controlled. However, in rare cases, it may progress to pleural fibrosis as the effusion resolves; this can limit reexpansion of the lung and results in varying degrees of pulmonary restriction and unexpandable lung (ie, trapped lung) [47,51,65]. Of importance, any surgical approach (eg, decortication) is not considered unless pulmonary restriction is still present after six months and limits the patient's exercise tolerance and quality of life.

Rarely, bronchopleural fistula formation and empyema necessitans (rupture of the empyema through the chest wall) can occur.

Pleural lymphoma associated with longstanding chronic pleural infection has been described particularly in Japanese patients [66,67].

PROGNOSIS — The long-term survival of patients with parapneumonic effusion and empyema is good, provided therapy is adequate and prompt.

In patients with pneumonia, the presence of a parapneumonic effusion is associated with an increased likelihood of being admitted and results in a longer hospital stay and increased mortality [68,69].

●Mortality is highest in those with empyema with retrospective series reporting a mortality of approximately 15 percent among empyema patients who are admitted to hospital [70,71]. In another prospective series of 85 patients, the mortality at four years was 14 percent with most deaths occurring in the first 400 days after drainage but many deaths were due to comorbid conditions or surgical complications rather than due to the empyema itself [18].

●One series reported that mortality was highest among those who require open surgery or decortication and 30-day readmission rates were highest in those who were managed with chest thoracostomy tubes alone (21 percent) compared with those treated with video-assisted thoracic surgery (VATS; 11 percent) and open surgery (13 percent) [1].

●In contrast, another retrospective analysis of over 9000 patients with a discharge diagnosis of empyema reported that over a 20-year period, patients treated with chest tube drainage with or without fibrinolytics had a higher mortality than those treated with VATS decortication (17 versus 11 percent); in addition there were no differences in readmission rates at 90 days [72].

●In a database analysis of over 21,000 hospitalizations for empyema, older age, number of comorbidities, were associated with increased odds of death [69].

Predictive scores such as the RAPID score has been described [73] and suggested by some guidelines groups for consideration in patients to inform decision-making [12].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pleural effusion".)

SUMMARY AND RECOMMENDATIONS

●Antibiotics – All patients with suspected (or diagnosed) parapneumonic effusion or empyema should be treated with antibiotics. Antibiotic therapy should be administered promptly and not delayed for sampling or drainage procedures. (See 'Antibiotic therapy' above.)

•General principles – Empiric antibiotic selection varies based upon the site of acquisition (ie, community- versus hospital-acquired), severity of illness, local epidemiology, and patient risk factors for drug-resistant pathogens or infection with other specific organisms. For complicated parapneumonic effusions and empyema, empiric regimens should generally include antibiotics that target anaerobes (table 1) and other likely pathogens (eg, streptococci if community-acquired; methicillin-resistant Staphylococcus aureus [MRSA] and Pseudomonas if hospital-acquired).

•Community-acquired – For infections in the pleural space that are acquired in the community, we suggest use of a third-generation cephalosporin (eg, ceftriaxone or cefotaxime [3^rd]) plus metronidazole or single agent therapy with a beta-lactam/beta-lactamase inhibitor combination (eg, ampicillin-sulbactam) (Grade 2C). For those who are penicillin allergic who cannot tolerate cephalosporins, we use single agent therapy with a carbapenem (eg, imipenem, meropenem) or combine metronidazole with either a fluoroquinolone (eg, levofloxacin) or a monobactam (eg, aztreonam).

•Hospital-acquired – For hospital-acquired infections (eg, health care-associated pneumonia or postprocedural empyema), we suggest combining vancomycin with metronidazole and an antipseudomonal cephalosporin (eg, cefepime, ceftazidime) or combining vancomycin with a beta-lactam/beta-lactamase inhibitor (eg, piperacillin-tazobactam, ticarcillin-clavulanate [limited supply]) (Grade 2C). For those who are penicillin-allergic who cannot tolerate cephalosporins, we suggest combining vancomycin with metronidazole and an antipseudomonal quinolone (eg, ciprofloxacin); alternatively, combining vancomycin with an antipseudomonal carbapenem (eg, imipenem or meropenem) is appropriate.

•Duration – The duration of antibiotic therapy for parapneumonic effusion and empyema is individualized. For most patients, we suggest continuing antibiotic therapy until there is clinical and radiographic improvement. In general, for self-resolving parapneumonic effusions, therapy may last one to two weeks, while therapy for complicated parapneumonic effusions and empyema are often longer (eg, two to three weeks for a complicated parapneumonic effusion and four to six weeks for empyema).

●Drainage – The following is a reasonable approach to drainage:

•Uncomplicated parapneumonic effusion – Patients with uncomplicated parapneumonic effusions (ie, typically small or moderate-sized, free-flowing effusion (table 4)), do not generally require drainage and are treated with antibiotics alone (algorithm 2 and algorithm 1). This approach is based upon the rationale that most of these effusions do not have direct microorganism invasion of the pleural space and resolve with antibiotics alone. Should patients or the effusion not improve, repeat imaging with chest computed tomography (CT) should be performed to evaluate for the development of a complicated parapneumonic effusion that may need to undergo sampling and drainage. (See 'Uncomplicated parapneumonic effusion (antibiotics alone)' above.)

•Complicated parapneumonic effusion – For patients with a complicated parapneumonic effusion, we suggest prompt drainage of pleural fluid in addition to antibiotic therapy (Grade 2C). A complicated parapneumonic effusion is one that has Gram stain, microbiologic or biochemical evidence of infection (algorithm 2). Although high quality data are lacking to guide selection of patients for drainage, such patients appear to be at risk for poorer outcomes including an increased requirement for more than one procedure, eventual need for surgery, and longer hospitalization. For patients with empyema, we recommend antibiotics and prompt drainage together with antibiotic treatment (Grade 1C), based upon the rationale that among those with parapneumonic effusion, those with empyema have the highest mortality. (See 'Complicated pleural effusion and empyema (antibiotics plus drainage)' above.)

•Tube type – The initial procedure of choice is a single tube or catheter thoracostomy. Image-guided placement of a catheter(s) may be needed when pleural loculations prevent adequate drainage by a single tube. During the next 24 to 48 hours, patients should be followed clinically (signs and symptoms, laboratory values, and drainage volume) and radiologically with chest CT to assess the response to drainage (degree of evacuation and reexpansion). (See 'Initial drainage (tube or catheter thoracostomy)' above and 'Assessment of response' above.)

•Duration – For those who clinically improve and in whom imaging reveals significant improvement or resolution of the effusion with adequate reexpansion of the lung, and the drainage rate has fallen to below 50 to 100 mL/day for two to three days, the chest tube or catheter can be removed, antibiotics continued, and the patient can be followed up as an outpatient in two weeks with repeat imaging. (See 'Discontinuing drainage' above.)

•Follow-up and reassessment – For patients who fail antibiotics and thoracostomy drainage from a well-placed chest tube (eg, the effusion persists or worsens, fever persists or new fever develops, persistent or worsening leukocytosis), attention should be paid to reassessing antibiotic coverage and need for additional drainage options. (See 'Treatment failure' above.)

-Ongoing or worsening infection – For patients with evidence of ongoing or worsening infection, options include reculturing pleural fluid directly from the pleural space or undrained locule (not from the tube or catheter) and adjustment of antibiotic coverage ensuring adequate anaerobic coverage or coverage of resistant organisms. Consideration should also be given to the presence of underlying necrotic lung as a focus of infection. (See 'Assess adequate antibiotic coverage and drainage' above.)

-Additional drainage options – For patients who require additional drainage, options include nonsurgical approaches using intrapleural tissue plasminogen activator/deoxyribonuclease (tPA/DNase) and/or placement of an additional drainage tube(s) or surgery, typically video-assisted thoracoscopic surgery (VATS). Choosing among the options is often provider-specific, and recommendations vary among clinicians, institutions, and guideline committees. Factors that influence the decision include the clinician's specialty, expertise availability, patient values, patient prognosis from comorbidities, and candidacy for select surgical procedures as well as the number and size of locules and degree of pleural thickening on chest CT. (See 'Choosing among drainage options' above.)

-tPA plus DNase – As an initial strategy, we suggest intrapleural administration of a combination of tPA 10 mg and DNase 5 mg, twice daily for three days rather than surgery (Grade 2C). The rationale for this approach is based upon data that report lower rates of referral for surgery (by 30 to 80 percent) when tPA/DNase are used. Drainage of residual locules is also appropriate, especially in those who fail tPA/DNase therapy. A repeat chest CT should be performed 24 to 48 hours after the chosen intervention(s) to evaluate the response. (See 'Assess adequate antibiotic coverage and drainage' above and 'Intrapleural tPA +/- additional drainage' above and 'Video-assisted thoracic surgery (VATS)' above.)

-Failed tPA/DNase – Patients who fail tPA/DNase and/or additional drainage procedures should be evaluated for VATS. (See 'Video-assisted thoracic surgery (VATS)' above.)

•Refractory patients – In patients refractory to nonsurgical therapy who are not candidates for or decline VATS, options are limited but may include persisting with conservative care (antibiotics and multiple drains) and in some cases, an open-window thoracostomy can be performed. (See 'Refractory cases' above.)

●Special populations – Patients with an infected pleural space following resective thoracic surgery (ie, lobectomy or pneumonectomy) require special attention since infection is often associated with a bronchopleural fistula that may also require surgical or endoscopic closure. Patients with malignant pleural effusion and an indwelling catheter who have an infected pleural space should be treated conservatively with continued indwelling catheter drainage and a prolonged course of antibiotics. (See 'Special populations' above and "Sequelae and complications of pneumonectomy", section on 'Pleural space complications' and "Bronchopleural fistula in adults".)

●Complications – Apart from complications associated with medical and surgical treatments, long-term complications of infection in the pleural space include residual pleural thickening on chest CT and rarely, when untreated, fibrothorax and pleural calcification. The pleural changes usually resolve over three to six months, and patients should not be considered for decortication unless they fail to improve and are symptomatic after six months. Rare complications include bronchopleural fistula formation, empyema necessitans (rupture of the empyema through to the chest wall), and pleural lymphoma. (See 'Complications' above.)

●Prognosis – The long-term survival of patients with parapneumonic effusion and empyema is good, provided therapy is adequate and prompt. Mortality is highest in those with empyema as well as in those who require open surgery or decortication. Death is highest during the first 400 days and is often due to comorbid conditions or surgical complications rather than due to the empyema itself. (See 'Prognosis' above.)