Ventilator-associated pneumonia in critically ill patients with COVID-19

Pandemic COVID-19 is associated with a high number of patients suffering from severe acute respiratory syndrome (SARS). Such patients can spend significant periods of time in intensive care units (ICU), with up to 80% of patients admitted to ICU requiring invasive mechanical ventilation [1, 2]. Critically ill patients are at high risk of nosocomial pneumonia, especially when ventilated [3]. The reasons for this includes breach of natural defences by invasive devices [4], sedation and impairment of coughing and mucociliary clearance, and the immunoparetic effects of critical illness [5, 6]. Early reports indicated that critically ill patients infected with SARS-CoV-2 had a high prevalence of nosocomial pneumonia, especially ventilator-associated pneumonia (VAP) [7]. More recent reports, including a large survey from a single hospital [8] and a synthesis of the literature [9] suggested that rates of secondary infections were low, although neither study focussed specifically on critical care. There remains considerable uncertainty around the incidence of nosocomial infections in severe COVID-19, which has led to recent calls for more analysis on the frequency, timing, and causative organisms of these important adverse events [10].

Reports of ICU-acquired infection in patients with COVID-19 have been limited and have often not reported the details of the causative organisms [7], or have focused on the incidence of one particular infection such as invasive aspergillosis [11]. Importantly, we are not aware of reports of ICU-acquired infections comparing patients with COVID-19 and those without managed contemporaneously within the same settings, which is key to interpreting the frequency, timing, and causative organisms leading to these infections.

Ventilator-associated pneumonia (VAP), the commonest ICU-acquired infection [3], can be challenging to diagnose as a range of non-infectious diseases may mimic the clinical picture of radiographic infiltrates, systemic inflammation and impaired oxygenation that typifies VAP [12]. To limit overdiagnosis and facilitate appropriate antimicrobial therapy in VAP, guidelines advocate the use of culture- based approaches [13, 14]. However, molecular tests to detect multiple pathogens (viruses and bacteria) are becoming more accessible and may further reduce unnecessary antimicrobial therapy [15] whilst enhancing the detection of hard to culture organisms.

During our hospital’s first wave of COVID-19 admissions we noted an apparent increase in the rate of VAP. In this study, we therefore aimed to identify and compare the distribution of VAP in critically ill ventilated COVID-19 patients compared to ventilated non-SARS-CoV-2 infected patients admitted to the same unit. We performed conventional microbiological culture on all lower respiratory tract samples. Bronchoalveolar lavage (BAL) was also analysed using a multi-pathogen TaqMan array card (TAC) we have developed and reported previously [16]. In a sub-set of BALs we assessed the composition of the bacterial lung microbiome in bronchoalveolar lavage (BAL) by 16S sequencing.

Overall, we managed 94 patients with COVID-19, of whom 81 were ventilated for more than 48 h. From the period 15th March to 30th August we also managed 144 patients without COVID-19 in the liver/general unit who required ventilation for more than 48 h. The demographic and clinical features of these two groups are shown in Table 1 and details of non-COVID admission diagnoses in Additional file 1: Table S1. Ventilator bundle audit data demonstrated high compliance (compliance with the full bundle ranged from 85 to 100%, with 99–100% for the period April–May when most COVID-19 patients were admitted).

Patients with COVID-19 were significantly more likely to be investigated for VAP (Table 1), and had a higher incidence of microbiologically confirmed VAP (39 (48%) COVID-19 patients compared to 19 (13%) patients without COVID-19). Further details of the comparison of the investigation for VAP are shown in Additional file 1: Tables S2 and S3. Patients who were investigated for VAP demonstrated a significant deterioration in oxygenation relative to the period immediately prior to the diagnosis (Additional file 1: Figure S1).

Survival analysis (Fig. 2) demonstrated that the increased risk of developing VAP in patients with COVID-19 was not simply a function of longer duration of ventilation. The hazard of early VAP was similar in both groups of patients, however the greater number of later VAPs in COVID-19 led to the increased median duration of ventilation before VAP developed seen in Additional file 1: Table S2. The effect of COVID status on VAP-free survival remained significant when adjusted for age and immunocompromised status (adjusted p value 0.045 by Cox proportional hazards model, Additional file 1: Table S5). Sensitivity analysis of patients with > 72 h mechanical ventilation and > 144 h of mechanical ventilation produced similar survival curves and hazard ratios (Additional file 1: Figure S2A and B). A similar finding was apparent when comparing crude incident density, patients with COVID-19 developed VAP at a rate of 28/1000 ventilator days, whilst those without COVID-19 experienced VAP at a rate of 13/1000 ventilator days (p = 0.009 by mid-P exact test). Incident density censoring for post-VAP duration ventilation, which is confounded by VAP itself prolonging ventilation, shows a similar pattern (40/1000 ventilator days for COVID-19, 19/1000 ventilator days for non-COVID p = 0.004 by mid-P exact test). Further details on timing of VAP are available in the supplemental section (Additional file 1: Table S2). Antibiotic use on admission (Table 1) and in the period leading up to investigation for suspected VAP (Additional file 1: Tables S3 and S4) was similar in frequency and spectrum of agents used.

The organisms identified on endotracheal aspirate culture and both culture and molecular testing of bronchoalveolar lavage fluid are show in Table 2. The concordance between culture and molecular testing was high, although molecular testing identified a number of additional organisms.

The distribution of organisms in COVID-19 and non-COVID-19 associated VAP is shown in Fig. 3, and is broadly similar between both groups.

COVID-19 is a new disease in the human population and this has led to an increase in the number of patients in need of mechanical ventilation, which in turn introduces the risk of VAP. COVID-19 can present in various severe manifestations and reports of co-infections vary [7‐9]. However, often these reports suffer from a lack of clarity around the severity of illness, location of patients (critical care vs non-critical care), timing of sampling relative to onset of disease and, where applicable, the use of mechanical ventilation. Here, we report on the most severely affected COVID-19 patients who required ICU admission with mechanical ventilation. We found that relative to patients without COVID-19, the hazard of VAP was significantly elevated.

Conventional surveillance for VAP uses incidence density, which we calculated to allow comparison with previously published reports. We found a high incidence density of confirmed VAP (28/1000 ventilator days) amongst patients with COVID-19, whilst those without COVID-19 had rates closer to those reported from other units in the pre-COVID-19 era, where incident densities were 6–14/1000 ventilator days for confirmed VAP were reported [23]. As sessional use of personal protective equipment remained in place until the end of July 2020, for management of both COVID-19 and non-COVID-19 patients, we do not think that this influenced the differential acquisition of VAP amongst these two groups.

The distribution of infecting organism was similar between patients with and without COVID-19, and reflects that reported in the literature from previous surveys of ICU-acquired infections from before the COVID era [3, 19]. The use of the TAC allowed more rapid identification of organisms, most of which were subsequently identified by culture. Notably there were a few organisms the TAC did not detect, largely because sequences for these organisms were not present on the card, these were distributed between both the COVID-19 and non-COVID patients.

At the lung microbiome level, we observed no difference in the composition of organisms between COVID-19 positive and non-COVID patients who developed VAP. Reassuringly, antibiotic susceptibility of the causative pathogens was similar in the two groups (data not shown) and this meant that conventional antimicrobial regimens could be used.

There is increasing recognition of fungal infections amongst patients with viral pneumonitides and VAP [11, 21, 22]. Although debate continues regarding the differences and similarities between influenza and COVID-associated aspergillosis [10], in keeping with our findings in bacterial VAP it appears that IPA is more common in COVID-19 patients than in ICU patients without COVID-19. It has been suggested that CAPA may relate to the use of immunosuppressive medications [10]. As can be seen from Table 1, steroids were relatively rarely used in this cohort of COVID-19 patients who were largely admitted before the results of the RECOVERY trial had been announced [24] and indeed none of the 3 CAPA patients we identified had received steroids prior to their diagnosis or had underlying immunosuppressive conditions.

More broadly, in our setting immunomodulatory medications were not commonly used at the time of the peak of the COVID-19 admissions, yet there remains a high prevalence of bacterial VAP in these patients. Although VAP in COVID-19 may present problems of quantity, we did not find evidence in this report of a qualitative difference in terms of the organisms causing infection, although as noted above aspergillosis may be more common although this needs to be seen in the context of a significantly higher rate of VAP overall. In the subset where we undertook microbiome profiling, our patients demonstrated similar profiles to those reported by other groups investigating the pulmonary microbiome of ventilated patients [25, 26]. The factors which lead to pulmonary dysbiosis in critical illness remain incompletely understood, but may include intercurrent antibiotic use, enteric translocation, pulmonary immune dysfunction and altered clearance [27].

Although patients without COVID-19 developed proportionately more ‘early’ VAP, being VAP within the first 4 days of ventilation (Additional file 1: Table S2), examination of the VAP-free survival curves (Fig. 2) reveals the hazard of early VAP is similar between the two groups. What is striking, however, is the ongoing risk of VAP seen in patients with COVID-19 which is greater than that seen in patients without COVID. This ongoing risk is reminiscent of the effect we reported previously in critically ill patients with marked immunoparesis [5].

Although from our observational study we cannot be certain why ventilated patients with COVID-19 have such a significantly increased risk of infection, previous work has indicated that the strongest predictor of nosocomial infection in critically ill patients is impaired immune cell function [5, 28]. Patients with COVID-19, in keeping with other critical illness syndromes such as bacterial sepsis and major trauma, experience a complex dysregulation of their immune function with features of both hyperinflammatory activation and organ damage as well as impaired antimicrobial functions [6, 29]. Notably, one of the key drivers of neutrophil impairment in critical illness is the complement component C5a [30, 31] and high levels of complement activation and C5a release have been reported in COVID-19 [32]. Other recent reports on the immunology of COVID-19 highlight marked increases in markers of immune cell functional suppression in the most severely unwell patients [29]. Damage to the alveolar membrane, although not specific to COVID-19, may also facilitate invasion of bacterial species [33]. The estimates of the prevalence of invasive pulmonary aspergillosis and herpesvirade reactivation are limited to those patients investigated by broncho-alveolar lavage and therefore represent only a subset of those investigated for VAP. In the case of invasive aspergillosis it also required senior clinician ordering of galactomannan, and as a retrospective study we cannot be sure clinicians had a common threshold for requesting this test. We therefore acknowledge that these data may underestimate the prevalence of these conditions, however the trend towards higher prevalence amongst patients with COVID-19 adds some support to the hypothesis that these patients suffer from a considerable burden of immunoparesis. It is notable, if not surprising, that patients with COVID-19 were much more likely to present with acute respiratory distress syndrome (ARDS) (Table 1) and consequently had more severe oxygenation defects and were much more likely to be ventilated prone. ARDS is an established risk factor for VAP [34], and the intense pulmonary inflammation can lead to immunologic reprogramming which can impair anti-microbial responses [35]. Prone positioning may increase risk of microaspiration, however dissecting out the specific effects of proning as opposed to the severity of the underlying lung inflammation remains challenging [36]. Whilst previous broad-spectrum antimicrobial therapy is an acknowledged risk factor for VAP [15] we did not find evidence of substantial differences in either antibmicrobial use or spectrum in patients with and without COVID19.

VAP remains difficult to definitively confirm without histological confirmation, which is seldom practical nor desirable in ventilated patients, we therefore cannot be certain that the patients with positive microbiology had definite pneumonia, although the use of quantitative cultures reduces the risk of detection of colonisation as opposed to infection [17, 37]. We used a clinically relevant definition similar to that used in previous studies [17, 37] and applied this consistently across the two groups. We note that diagnostic technique can alter the rate of diagnosis [37], and therefore think it is reassuring that the proportion of bronchoscopic diagnoses was consistent across both groups (Table 1). Similarly, the use of the more sensitive TAC molecular diagnostic could increase the apparent rate of microbiologically confirmed VAP, it is reassuring that TAC was used marginally more frequently in the non-COVID patients and this also suggests the difference seen is due to biological rather than technical reasons.

Reports of rates of VAP amongst ventilated patients with COVID-19 vary, with rates of 40–86% reported [38‐40] and our reported rate of 49% is in keeping with reports from other centres. Although some reports, not focussed specifically on VAP, indicate lower rates of 10% [8], it is unclear how many of the ICU patients in that cohort were ventilated for at least 48 h. The rates of VAP between centres managing COVID-19 are likely to vary depending on the clinical characteristics of the patients managed, differential ICU admissions policies and clinical factors such as use of immunosuppressive therapies. Although we managed to maintain one to one nursing ratios throughout the first wave of COVID-19, it is possible that the increased numbers of nurses with only brief training in critical care led to increased rates of VAP. However, the continued high compliance with the ventilator care bundle which includes key nursing interventions such as oral hygiene and head of bed elevation argues against this being a major factor. We acknowledge the sample size and single centre limitations with our observations and suggest larger studies from distinct geographic locations may help fully understand the risk of developing secondary bacterial infections in patients with severe COVID-19.

COVID-19 makes people more susceptible to developing VAP, partly but not entirely due to the increased duration of ventilation. The change in lung microbiome and causes of secondary infection are similar to those seen in critically ill patients ventilated for other reasons. Careful sampling of the respiratory tract whilst minimising contamination from the proximal tract, in combination with sensitive diagnostic testing to reduce the risk of false negative cultures will aid antimicrobial optimisation in patients with COVID-19.