My approach to respiratory sepsis: pulmonary

Abstract

Respiratory sepsis is one of the leading causes of sepsis-related mortality worldwide, with significant impact on the healthcare burden in hospitals. While there are standardised guidelines for the management of sepsis in general, we aim to specifically review the approach to respiratory sepsis. This review discusses common aetiologies and the pathophysiology of respiratory sepsis, as well as aspects of management such as oxygenation and ventilation, methods of source identification, airway secretion management, and the role of glucocorticoids.

Shareable abstract

Respiratory sepsis is a leading cause of sepsis-related mortality. This article reviews the aetiology and pathophysiology of respiratory sepsis, as well as aspects of management such as oxygenation, source identification and the role of physiotherapy. https://bit.ly/4prKwnv

Educational aims

• To understand the pathophysiology and pathogens of respiratory sepsis.

• To describe modes of respiratory support involving supplemental oxygen delivery and ventilation support such as high-flow nasal cannula and noninvasive ventilation.

• To review the role of traditional sputum collection, induced sputum and bronchoscopy in respiratory sepsis.

• To understand the role of chest physiotherapy in the management of respiratory sepsis and the criteria to consider referral.

Introduction

Sepsis: Life-threatening organ dysfunction as a result of dysregulated host response to infection, as per the Third International Consensus Definition for Sepsis and Septic Shock (Sepsis-3) [1].

In 2017, a global study of sepsis incidence and mortality attributed lower respiratory tract infections as the second leading cause of sepsis incidence worldwide, and the most common underlying cause for sepsis-related death [2]. In the Western world, lower respiratory tract infections are the leading cause of sepsis, accounting for up to half of all cases [3]. We define respiratory sepsis as the presence of organ dysfunction as defined by the Sepsis-related Organ Failure Assessment score in the setting of an established respiratory tract infection.

Respiratory sepsis is the presence of a sepsis-related organ failure assessment score of two or greater in the setting of a respiratory tract infection (established or suspected).

Establishing the exact rate of respiratory sepsis-related mortality is a challenge due to the advancement of healthcare infrastructures promoting both increased and earlier detection and treatment of sepsis in recent years. Several studies have reported that patients who develop severe respiratory sepsis have a mortality rate of 20–30% [4–6].

When we discuss sepsis in relation to the lungs, it is essential to acknowledge the historical concepts of acute lung injury and acute respiratory distress syndrome (ARDS), as they are invariably the result of severe respiratory sepsis. It is important to acknowledge that the criteria for ARDS have changed over the past two decades, which is highlighted in table 1 [7–9].

TABLE 1.

Criteria for acute respiratory distress syndrome (ARDS) based on the American–European Consensus Conference (AECC) definition (1994), the Berlin Definition (2012) and the global ARDS definition (2023)

	AECC (1994) [7]	Berlin Definition (2012) [8]	Global ARDS (2023) [9]
Onset	Within 1 week of a known clinical insult	Within 1 week of a known clinical insult or new/worsening symptoms	Similar to Berlin Definition
Imaging	Bilateral infiltrates on chest radiograph	Bilateral opacities on chest radiograph or CT, not fully explained by effusion, collapse or nodules	Bilateral opacities on chest radiograph or lung ultrasound
Pulmonary oedema	Noncardiogenic (no evidence of left atrial hypertension)	Not fully explained by cardiac failure or fluid overload; echocardiography if needed	Same as Berlin Definition; clinical judgement or ultrasound used
Oxygenation	ALI: PaO2/FIO2 ratio 200–300 mmHg ARDS: PaO2/FIO2 ratio ≤200 mmHg	Mild: PaO2/FIO2 ratio 200–300 mmHg Moderate: PaO2/FIO2 ratio 100–200 mmHg Severe: PaO2/FIO2 ratio ≤100 mmHg	Same as Berlin Definition; SpO2/FIO2 ≤315 when PaO2 not available (using pulse oximetry)
PEEP	Not required	Minimum PEEP of 5 cmH2O required	No PEEP requirement due to limited ventilation resources
Other considerations	ALI term included	Removed ALI term; clarified severity categories and criteria	Designed for low-resource settings; uses SpO2/FIO2 and ultrasound as substitutes

PEEP: positive end-expiratory pressure; CT: computed tomography; ALI: acute lung injury; P_aO2: arterial oxygen tension; F_IO2: inspiratory oxygen fraction; S_pO2: peripheral oxygen saturation.

Pathophysiology

As stated in the definition of sepsis, the core driver of sepsis (of any cause) is the dysregulated host immune response to a pathogen. The pathophysiology of this response is complex, encompassing multiple different processes. The commonly proposed theory is the activation and excessive production of pro-inflammatory cytokines, initially aimed to combat infection, which leads to cytotoxic effects and significant inflammation [4, 10]. Patients with severe sepsis also develop a variable degree of immunosuppression due to hypofunctional neutrophils and lymphocyte depletion [10]. The combination of these factors results in an imbalance in immune function and, subsequently, progression of sepsis.

The alveolar epithelial cells lining the alveolar wall provide protection to the lung from infections in normal physiological conditions by promoting alveolar fluid clearance, maintaining surface tension within the alveoli, and decreasing pulmonary inflammation [11]. In the setting of sepsis and excess inflammation, the destruction of alveolar epithelial cells by pro-inflammatory cytokines can be seen as the driver of sepsis-related lung injury.

Monocytes are one of the main components in the body's innate immune response that are produced in the bone marrow and migrate to the site of infection or inflammation where they differentiate into activated macrophages [12, 13]. They can differentiate into either M1 macrophages, which mediate destruction of pathogens via inflammation, or M2 macrophages, which promote tissue repair and clearance of apoptotic cells. This is an important process in the setting of respiratory sepsis, where the body is unable to effectively mount an adequately balanced response to pathogens within the lungs, the proinflammatory process will go unmitigated. This is due to the relatively unopposed activity of M1 macrophages releasing cytokines such as tumour necrosis factor-α, interleukin (IL)-1, IL-6 and IL-8, which exacerbate inflammation.

Aetiology

While sepsis is a result of infection, finding a causative pathogen is often difficult. A study of 6.8 million patients with severe sepsis in the USA showed that roughly half of the patients had negative blood and/or respiratory cultures [14]. When focusing on respiratory sepsis, the question is whether it is driven by bacterial, viral or fungal infections. The exact epidemiology of each is highly dependent on the population demographics and pathogens endemic to the area. The majority of cases will be due to viruses or bacteria, with fungal causes predominantly affecting immunocompromised hosts.

Based on multiple studies, bacterial and viral infections are the most common causes of respiratory sepsis. In a study of 2488 patients in the USA admitted with community-acquired pneumonia, 38% had a pathogen identified, with more than half of the cases being due to a virus [15]. A study in Spain detected a causative pathogen in 42% of patients with community-acquired pneumonia (14% treated in the outpatient setting) and conversely showed that 90% of cases in their centre were due to bacterial causes [16]. Bacterial species linked with more severe pneumonia and sepsis were identified as Gram-negative enteric bacilli, Pseudomonas spp. and Staphylococcus aureus. Viral respiratory sepsis tends to follow a more seasonal pattern, with influenza A and influenza B viruses being the culprit for most cases, particularly during the winter months [6, 17]. Blood culture plays an important role, as it may determine an infective source and aid source control in pneumonia. Other common viruses include rhinovirus, parainfluenza virus, coronavirus, human metapneumovirus and respiratory syncytial virus (RSV). There is the suggestion that viral pneumonias that potentially lead to respiratory sepsis are underdiagnosed, likely due to limitations in the testing facilities and capabilities of individual laboratories [17].

Bacterial and viral pneumonias can be distinguished by the detection of causative bacteria by culture or serological testing. Bacterial pneumonias are associated with a markedly elevated blood procalcitonin, C-reactive protein or leukocyte levels compared to viral pneumonias. Studies of pneumonia in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus infection highlighted haematological abnormalities such as lymphocytopenia and monocytosis as indicators [18].

It is important to note the importance of viral and bacterial co-infection, as studies have reported that influenza virus and RSV co-infection with bacteria such as Streptococcus pneumoniae or Staphylococcus aureus have resulted in more severe disease, with a third of fatal influenza cases identifying a bacterial co-infection [19]. Research has shown that respiratory viruses are able to damage the respiratory epithelium, which may contribute to bacterial adherence to the basement membrane. Increased neutrophil apoptosis has been observed in influenza A infections, which may dampen bacterial clearance [20].

Source identification

Identifying the pathogen responsible in cases of respiratory sepsis is an important aspect of the pulmonologist's approach, particularly when considering that most cases do not have a pathogen identified. The main method of identifying a source specific to the respiratory system is through sputum cultures. The diagnostic yield of expectorated sputum cultures is highly variable, with up to 50% of patients with pneumonia not able to produce sputum in a timely manner, prior to administration of antibiotics [21]. As mentioned earlier, induced sputum is a consideration. Due to the risk of bronchoconstriction with nebulised hypertonic saline, patients should have close monitoring of their forced expiratory volume in 1 s, particularly if they have asthma or severely impaired lung function [22]. The limitation of induced sputum is that it requires trained staff and close monitoring of the patients during the procedure, which can be technically demanding. Studies have been performed comparing the diagnostic yield of induced sputum with expectorated sputum, with results showing no significant difference between them [23, 24]. While there may not be a diagnostic advantage over expectorated sputum, induced sputum might be useful in cases where the patient is unable to expectorate sputum on their own, although the time and staffing required will be a limitation. Overall, induced sputum does not have a significant advantage over expectorated sputum and should not be routinely performed in the evaluation of respiratory sepsis, given the risk of bronchoconstriction and requirement of trained staff.

To assess for viral pathogens, a nasopharyngeal swab is usually taken and analysed with PCR testing for common respiratory viruses such as influenza A, influenza B, RSV and SARS-CoV-2. The role of an extended respiratory virus panel can be considered in cases of slowly resolving lower respiratory tract infections following discussion with microbiology, although this is usually not performed in the acute setting of pulmonary sepsis.

Urinary antigens for Streptococcus pneumoniae and Legionella pneumophila pneumonia are commonly used in the diagnostic workup of pneumonia. These tests can provide rapid identification of the pathogen, although the 2019 American Thoracic Society (ATS)/Infectious Diseases Society of America (IDSA) guidelines only have a weak recommendation for the routine use of these tests, as randomised trials did not show any outcome-related benefits [25, 26]. In addition to this, studies assessing the overall cost-effectiveness of urinary antigen testing suggests that the cost per positive result does not offer a benefit when compared with potential savings made through antimicrobial changes and that it had no impact on admission durations [27, 28]. This can be because the incidence of Legionella pneumonia, while severe, is low within the community and results in few positive results, even in cases with high suspicion such as recent travel, hyponatraemia or diarrhoea. Conversely, Streptococcus pneumoniae is one of the most common pathogens leading to pneumonia and its detection in urinary antigen testing does not necessarily change management, as it would already be treated by traditional broad-spectrum antibiotics. Considering the evidence base behind this, we would agree with the ATS/IDSA guidelines that urinary antigens should only be tested in cases of severe pneumonia, in which respiratory sepsis would be included, with high-risk clinical features such as hyponatraemia, travel history, or gastrointestinal symptoms.

Fibreoptic bronchoscopy is another method of source identification in pulmonary sepsis. According to the British Thoracic Society guidelines, bronchoscopy should be considered in patients with nonresolving or slowly resolving pneumonia despite antimicrobial therapy, particularly in patients aged >50 years with a significant smoking history [29]. It specifically states that the role of bronchoscopy in the acute phase of community-acquired pneumonia is limited, as noninvasive investigations are preferred, although no exact cut-off period is noted. Bronchoalveolar lavage (BAL) samples can have sensitivities up to 77%, which is the test with the highest diagnostic yield with respect to pulmonary infections. The main concern with performing bronchoscopy in an acutely unwell patient with pulmonary sepsis is the degree of hypoxia and the use of sedative medications, which can be technically challenging. A study regarding the safety of diagnostic bronchoscopy in patients with pulmonary sepsis reported it to be well tolerated, with 10% of cases being complicated by transient hypoxia [30]. This study is limited by a small sample size and the fact that bronchoscopies were carried out on mechanically ventilated patients in the intensive care unit, which does not reflect patients at ward level care. There have been studies that support the use of bronchoscopy in patients with unresolving pneumonia, showing an additional diagnostic value in ∼50% of patients with initial treatment failure and guiding antimicrobial changes, although it did not show mortality benefits, probably due to the severe nature of the infection [31, 32]. One study highlighted that the mean time between hospital admission and bronchoscopy was 14 days, which is in line with an extended period of initial antimicrobial therapy prior to determining treatment failure [33]. Overall, the evidence suggests that the role of bronchoscopy and BAL in patients with respiratory sepsis is limited in the initial phases, but should be considered when patients fail to respond to initial antimicrobial therapy where there is the suspicion of an atypical infection.

The approach to a patient with respiratory sepsis with the aforementioned investigations and interventions are summarised in figure 1.

FIGURE 1.

General approach to a patient with respiratory sepsis. FBC: full blood count; CRP: C-reactive protein; P_aO2: arterial oxygen tension; F_IO2: inspiratory oxygen fraction; ARDS: acute respiratory distress syndrome.

Imaging

The majority, if not all, of respiratory sepsis cases will be due to an underlying pneumonia. While we can suspect a pneumonia based on classical lower respiratory tract symptoms, the confirmation of pneumonia is done radiographically with either chest radiography or computed tomography (CT) imaging. In the incident where imaging may not show any evidence of pneumonia within the lung parenchyma, clinical judgement must be used, as lower respiratory tract symptoms in the absence of radiographic evidence of pneumonia could reflect the early stages of a developing pneumonia or an uncomplicated lower respiratory tract infection, although it is not common for it to progress to sepsis. Studies have reported that initial chest radiographs in patients with suspected pneumonia do not show consolidation in up to a third of cases [34, 35]. It has been proposed that up to 30% of lung lesions can be missed on a posteroanterior view chest radiograph [36]. This is mainly due to overlapping structures such as the diaphragm and solid organs that can conceal a proportion of lower lobe pneumonias. If there is a high degree of suspicion for a lower lobe pneumonia, lateral chest radiography can provide an unobstructed view of the lung bases. However, studies have noted that it has no role in the routine workup of pneumonia, mainly due to the higher dose of ionising radiation the patient is exposed to [37].

Furthermore, in patients with an inconclusive posteroanterior chest radiography, there is a higher likelihood of proceeding towards a CT thorax. A study that assessed the diagnostic value of chest radiography and CT thorax reported that 41% of pneumonia cases were diagnosed on CT following negative or inconclusive chest radiographs [38]. While this is a significant tool in securing a diagnosis, the risk of higher radiation exposure and increased burden on healthcare systems should be considered when establishing the need to proceed with a CT thorax.

Point-of-care lung ultrasound has become an emerging modality in recent years, with several studies reporting that lung ultrasound has a higher sensitivity and specificity in diagnosing pneumonia when compared to chest radiography [39–41]. Studies have shown that there are objective and reproducible features especially in children that indicate pneumonia on lung ultrasound, such as B-lines, air bronchograms and pleural effusions (if present) [40].

Based on the reviewed studies, a CT thorax should be performed in a patient presenting with signs and symptoms of pulmonary sepsis when the initial chest radiograph does not show an obvious consolidation to accurately determine the presence of lung infiltrates. Otherwise, alternative aetiologies should be investigated. It is important to consider that tachypnoea and hypoxia can be a result of tissue hypoxia driven by sepsis from a separate organ system. In rare cases where radiation exposure is avoided, magnetic resonance imaging can be considered, although its role is limited due to motion artefact and lower spatial resolution.

Considerations for a respiratory-based approach

Patients with respiratory sepsis often experience significant hypoxaemia. Broadly speaking, tissue hypoxia is a well-established complication of sepsis due to the inflammatory cytokines affecting the capillary walls leading to vasodilation and a reduction in systemic blood pressure, which in turn leads to reduced oxygen delivery to the body cells. Patients with sepsis can also develop ARDS, where the increased vascular permeability affects the pulmonary capillaries, resulting in fluid accumulation the alveolar spaces [42].

In addition, patients with pneumonia will have direct injury to the lung through a variety of mechanisms. The pathogen can induce cytotoxic effects on the alveolar epithelial barrier, with the inflammatory process leading to alveolar oedema, resulting in a ventilation and perfusion mismatch [43]. Mucus in the airways act as a defensive barrier to promote clearance of foreign substances such as microbes, inhaled pollutants, or inflammatory cells. In the setting of a respiratory infection, the inflammatory response will drive increased mucus production with the aim of helping to clear out microbes and white blood cells that are involved in the phagocytosis of pathogens. Damaged respiratory epithelium leads to a reduction in the ability of the airways to clear mucus. In severe infection, there can be an imbalance between mucus production and clearance, which results in significant amounts of mucus remaining in the airways and lead to mucus plugging. When these factors are combined, the patient will potentially suffer from severe hypoxaemia.

Considering the mechanisms mentioned so far, our discussion regarding the approach to respiratory sepsis will include the management of hypoxaemia with oxygen and ventilation, the management of airway secretions, and the methods of source identification involving induced sputum or bronchoscopy.

Oxygen and ventilation

Providing adequate oxygenation is one of the key features in the management of pulmonary sepsis. It is important to understand that hypoxaemia is defined by a low concentration of oxygen in arterial blood. This inevitably results in hypoxia, which is defined by the lack of adequate oxygen supply to the tissue. Supplemental oxygen is used to treat hypoxaemia and reverse tissue hypoxia. There are several different interfaces used to deliver supplemental oxygen to patients.

In emergency situations, nonrebreather masks are used to provide a high concentration of oxygen within a short period of time. They have a reservoir bag which is filled with oxygen at high flow rates of 12–15 L·min⁻¹ to prevent collapse of the bag. When patients are more stable with mild/moderate hypoxia, the nonrebreather mask is stepped down to typical oxygen delivery devices such as nasal cannula, face mask or Venturi masks.

If the patient remains persistently hypoxic and not able to maintain adequate oxygenation on the typical oxygen delivery devices, which is the case in the majority of respiratory sepsis cases, devices that deliver higher concentrations of oxygen should be considered. Two of the main interfaces used to deliver this include high-flow nasal cannula (HFNC) and noninvasive ventilation (NIV). These administer oxygen at higher flow rates than conventional oxygen delivery devices, providing a level of pressure support in the form of positive end-expiratory pressure (PEEP). The concept of PEEP has significant importance when discussing ventilatory support in patients with respiratory compromise and/or sepsis due to the positive expiratory pressure splinting the alveoli open, reducing atelectotrauma, thus increasing alveolar recruitment and oxygenation.

HFNC consists of an oxygen/air blender that provides a fraction of inspired oxygen (F_IO2) ranging from 21% to 100% that connects to an active heated humidifier which delivers a high flow of oxygen via nasal cannula, allowing titration of the oxygen concentration independent to the flow rate [44]. The humidified and warmed air mimics physiological conditions that allows the high flow rate to be better tolerated by patients, and reduces the work of breathing in patients suffering from respiratory sepsis [45]. Studies have reported that the high flow provides a low level of positive airway pressure, equivalent to a PEEP of 2–4 cmH₂O [46, 47]. The use of HFNC was widely studied during the COVID-19 pandemic, and studies during this period have suggested that there is not a higher risk of aerosolisation when compared to standard oxygen masks [48].

Typically, ventilatory support is provided as either continuous positive airway pressure (CPAP) or bi-level positive airway pressure (BiPAP). For the purposes of simplicity in this article, we use the umbrella term, NIV. As the name implies, CPAP provides continuous pressure support, whereas BiPAP provides two levels of pressure, enabling ventilatory support. BiPAP is separated into inspiratory positive airway pressure and expiratory positive airway pressure, the latter of which corresponds to PEEP. The interface for oxygen delivery differs from HFNC, with closely fitting masks covering the mouth and nose to provide a targeted level of PEEP. CPAP is indicated in the treatment of hypoxaemic respiratory failure and/or pulmonary oedema, while BiPAP is traditionally used in hypercapnic respiratory failure with acidosis. A recent systematic review of BiPAP use in hypercapnic respiratory failure in patients with COPD showed that it reduced the risk of mortality by 46% and reduced the progression to mechanical ventilation by 65%, although it was noted to have a higher treatment intolerance rate compared to conventional oxygen therapy [49]. The key features of each oxygen delivery device are highlighted in table 2.

TABLE 2.

Key features of different oxygen delivery devices

	Flow rate/FIO2	Mechanism/interface	Positive pressure/PEEP	Indications	Considerations/notes
Nonrebreather mask	12–15 L·min−1; high FIO2 (up to ∼90–100%)	Reservoir bag+one-way valves; face mask	No PEEP, but reservoir prevents room air entrainment	Emergency/high hypoxia needing rapid oxygen boost	Risk of suffocation if oxygen supply fails; used acutely, then stepped down when stable
Conventional oxygen devices	Nasal cannula, simple face mask, Venturi mask	Low flow (up to 6 L·min−1 nasal cannula); Venturi delivers fixed FIO2	No PEEP	Mild to moderate hypoxia	Low flow, no pressure support
High-flow nasal cannula	Up to 60 L·min−1; FIO2 21–100%	Heated, humidified oxygen via nasal cannula	Provides low PEEP (∼2–4 cmH2O), variable with mouth open	Persistent hypoxia despite conventional devices	Improves comfort and secretion clearance; variable PEEP; lower aerosol risk than noninvasive ventilation
Noninvasive ventilation (CPAP/BiPAP)	Flow depends on settings; FIO2 adjustable	Tight-fitting face mask covering nose and mouth	CPAP: continuous positive pressure; BiPAP: IPAP+EPAP (PEEP)	CPAP: hypoxaemic respiratory failure, pulmonary oedema BiPAP: hypercapnic respiratory failure with acidosis	Risk of lung overdistension if pressures too high; BiPAP reduces mortality and intubation risk in hypercapnic respiratory failure; higher intolerance rate; classified as aerosol-generating procedure

F_IO2: inspiratory oxygen fraction; PEEP: positive end-expiratory pressure; CPAP: continuous positive airway pressure; BiPAP: bilevel positive airway pressure; IPAP: inspiratory positive airway pressure; EPAP: expiratory positive airway pressure.

The Irish Thoracic Society guideline on oxygen therapy [50] suggests an initial assessment of oxygenation based on a pulse oximetry reading. In an emergency setting and/or with oxygen saturations <85%, a nonrebreather mask at 15 L·min⁻¹ should be used before more controlled oxygen therapy can be provided. If the patient is clinically stable with oxygen saturations >85%, nasal cannula with a flow rate ranging from 1 to 4 L·min⁻¹ can be used, with oxygen therapy escalated from that point. If the patient remains breathless or hypoxic while on 4 L·min⁻¹ of conventional oxygen therapy, HFNC at an initial flow rate of 30 L·min⁻¹ should be started, and the concentration of oxygen titrated to their target oxygen saturation level. The flow rate can be increased up to 60 L·min⁻¹ depending on the patient's tolerance and their oxygen requirements. The reason for the escalation to high-flow humidified oxygen is due to the prolonged duration of nonhumidified oxygen that can lead to drying of airway mucosa and the lack of support for work of breathing, which can be severe in patients with respiratory sepsis. If the patient is not responding to HFNC at maximally tolerated flow rate and F_IO2 concentration, the escalation to CPAP for more targeted pressure support should be considered and discussed with the patient and/or their family. CPAP can be started at a pressure of 5 cmH₂O and titrated upwards depending on the clinical response. Special considerations must be made in cases of respiratory sepsis with hypercapnic respiratory failure with acidosis, as BiPAP is the first-line intervention with controlled oxygen therapy.

A study comparing the use of HFNC with conventional low-flow oxygen therapy in patients with severe COVID-19 infection showed that HFNC had a lower rate of progression to mechanical ventilation and improved recovery times [51]. This supports the concept that HFNC is effective in reducing the work of breathing and improving oxygenation in patients with respiratory sepsis. More importantly, there have been several studies comparing HFNC with NIV in the treatment of hypoxaemic respiratory failure, which can be due to severe respiratory sepsis [52–57]. In the 2015 FLORALI study, HFNC demonstrated better 90-day mortality in acute hypoxaemic respiratory failure when compared to standard oxygen therapy and NIV, with similar intubation rates [53]. The implication of this study is that progression to NIV from HFNC may not be appropriate for all patients, particularly those with high frailty scores and multimorbidity as the outcomes are similar with NIV potentially causing more discomfort and distress to the patient. The recent RENOVATE randomised clinical trial reiterated that HFNC is noninferior to NIV in the management of acute hypoxaemic respiratory failure with regards to endotracheal intubation rates and 7-day mortality [57]. These results support the important role of HFNC in the management of respiratory sepsis.

Positive airway pressure therapies provide mechanical support for secretion management in select patients, i.e. those with excessive dynamic airway collapse. CPAP maintains constant airway pressure throughout the respiratory cycle, helping to splint open the small airways, prevent alveolar collapse, and enhance ventilation behind retained mucus [58, 59]. In more severe cases where patients demonstrate respiratory fatigue or hypercapnic respiratory failure, BiPAP may be preferred. BiPAP provides differential pressures (higher inspiratory and lower expiratory pressures), augmenting tidal volumes and enhancing cough flows needed for secretion clearance. Careful monitoring of haemodynamic status, oxygenation, and patient tolerance is essential during positive pressure therapy.

Physiotherapy management of secretion clearance in pulmonary sepsis

Pulmonary sepsis frequently leads to excessive, thickened airway secretions due to infection, inflammation and impaired mucociliary function. Effective secretion management is critical to optimise ventilation, reduce respiratory workload, and prevent secondary complications such as atelectasis and worsening hospital acquired pneumonia [58, 60]. The role of chest physiotherapy is essential in this regard, as their involvement with secretion management, airway clearance techniques, and early supported mobilisation aids in expediting patient recovery and improving respiratory status.

Referral to physiotherapy for secretion management

Medical teams should consider a referral to physiotherapy colleagues when patients with pulmonary sepsis exhibit clinical signs of secretion retention causing an increase in their work of breathing, including but not limited to coarse crackles, visible or audible mucous retention, weak or ineffective cough, and deterioration in gas exchange despite appropriate medical therapy [59, 60].

Other triggers for referral include new radiological findings (e.g. lobar collapse, mucus plugging). Those with underlying conditions such as COPD, bronchiectasis or neuromuscular weakness should warrant consideration for early referral [48, 51]. Early referral for physiotherapy intervention can prevent further respiratory decline and reduce the need for invasive ventilation [60, 61].

Role of physiotherapy in secretion management

Physiotherapists employ a range of airway clearance techniques to facilitate mucous clearance in pulmonary sepsis. Techniques such as the active cycle of breathing technique (ACBT), and use of oscillatory positive expiratory pressure devices, i.e. Aerobika, Flutter, Cornet, etc., aim to mobilise secretions from peripheral to central airways to promote secretion clearance [45, 53]. Manual techniques such as chest percussion and manual chest vibrations may also be used in select patients.

Assisted coughing techniques, i.e. in patients with neuromuscular compromise, suctioning (in patients unable to clear secretions independently), breathing exercises, postural drainage, and mobilisation are integral to comprehensive physiotherapy care [58–60].

Ambulation/mobilisation for secretion clearance

Mobilising patients as early as clinically feasible has been shown to significantly enhance secretion clearance. Moving promotes larger tidal volumes, improves diaphragmatic function, and encourages ventilation of dependent lung regions, facilitating collateral ventilation behind secretions [61, 62]. Gravity-assisted drainage is naturally enhanced in upright positions, aiding mucous movement toward the central airways [47].

Additionally, physical activity improves mucociliary clearance and strengthens cough efficacy, making secretion clearance more efficient [58, 62]. Mobilisation also improves overall circulation, reduces sedation-related complications, and supports better immune function [61, 62]. As a result, early and progressive mobilisation forms a cornerstone of physiotherapy management in pulmonary sepsis.

Obtaining sputum samples during physiotherapy

Physiotherapists may assist in obtaining sputum samples for microbiological analysis during airway clearance sessions. If a patient can expectorate during ACBT, huffing or manual techniques, sputum can be collected into a sterile container for testing [60, 63]. If expectoration is not possible, physiotherapists may perform suction-assisted sputum collection using sterile techniques and use a suction trap [59, 60]. In certain cases, sputum induction with nebulised hypertonic saline may be considered, provided appropriate protocols are in place [60, 64]. Collection should be halted if severe desaturation, respiratory distress, or haemoptysis occurs, and urgent medical review sought.

Pharmacotherapy

Local guidelines dictate appropriate empirical antimicrobial therapy, which varies according to geographical variation in prevailing community and hospital-based microbial patterns; previously detected microbial pathogens for an individual; allergy history; and severity of the acute illness.

The role of glucocorticoids in the management of pulmonary sepsis has become more prominent in recent times. A study in 1999 by Montón et al. [65] suggested that glucocorticoids decrease the inflammatory response systematically and locally within the lungs in patients with severe pneumonia. Subsequent randomised controlled trials did not consistently show a benefit of glucocorticoids on clinically relevant morbidity and mortality outcomes, although recent trial data have shown compelling evidence supporting the early administration of glucocorticoids in patients with pulmonary sepsis. In 2023, a multicentre trial of 800 patients with severe pneumonia showed that patients who received glucocorticoids had a lower 28-day mortality rate [66]. The REMAP-CAP Corticosteroid Domain Randomized Clinical Trial published in 2025 did not demonstrate a significant difference in mortality rates with the use of glucocorticoids in severe pneumonia, although this study had an unbalanced cohort of patients, with 536 receiving glucocorticoids and 122 in the control group [67]. The recent APROCCHSS trial showed that patients with severe pneumonia and septic shock who received fludrocortisone in addition to hydrocortisone had reduced mortality risk, it is worth noting that this finding was not seen in septic shock unrelated to pulmonary infection [68]. This evidence suggests that glucocorticoids can be beneficial in pulmonary sepsis by attenuating the inflammatory response within the lungs and should be a consideration in the management of the disease.

Conclusion

Pulmonary sepsis is a condition with significant morbidity and mortality, driven by an unmitigated host immune response to a lower respiratory tract infection. This review highlights important considerations in the evaluation of the condition, focusing on the radiological and microbiological investigations. Core issues that warrant specialist management from a pulmonary perspective include escalation of oxygen therapy with or without ventilatory support, and the approach to airway secretion clearance. Physiotherapy referral for secretion management in pulmonary sepsis reduces airway obstruction, improves ventilation, enhances oxygenation, and prevents further lung injury or infection, with the overall aim of expediting a patient's discharge. Recent data support the use of glucocorticoids in pulmonary sepsis. An evidence-based approach must be adopted to improve patient outcomes and establish the necessity of different interventions.

Key points

• HFNC provides respiratory support in the form of humidified oxygen at higher flow rates and is not inferior to NIV in preventing progression to intubation in patients with respiratory sepsis.

• Early referral to physiotherapy for airway clearance techniques is crucial in the management of pulmonary sepsis.

• Pathogen identification is important to guide antimicrobial therapy, but the method of sampling (expectorated sputum/induced sputum/bronchoscopy with BAL) depends on availability of sputum, local expertise regarding sputum induction and the initial response of the patient to initial antimicrobial treatment.

• Glucocorticoids have been shown in recent studies to have a beneficial role in the treatment of pulmonary sepsis and should be considered in the management process.

Self-evaluation questions

1. What technique promotes the clearance of mucus from peripheral airways to central airways?

   1. Active cycle of breathing techniques
   2. Suctioning
   3. Assisted coughing
   4. Early mobilisation
2. What is the role of bronchoscopy with BAL in the management of pulmonary sepsis?

   1. First-line investigation to determine causative pathogen
   2. Therapeutic for secretion clearance
   3. Considered as further investigation in slowly responding disease
   4. No clear role and should not be performed
3. What is the mortality rate in patients with severe respiratory sepsis?

   1. 10%
   2. 30%
   3. 50%
   4. 70%
4. How does HFNC provide support to patients in respiratory sepsis?

   1. Provides a pressure support to splint open the airway and improve oxygenation
   2. Recruitment of alveoli and reducing dead space
   3. Provides humidified oxygen to improve airway secretion clearance
   4. Provides ventilatory support to enhance cough support and secretion clearance

Suggested answers

1. a.

2. c.

3. b.

4. c.