Striking a Balance: Embracing Clinical and Surveillance Frameworks for Comprehensive Infection Surveillance

INTRODUCTION Ventilator-associated pneumonia (VAP) is one of the most prevalent healthcare-associated infections and has a significant impact on patient outcomes and associated healthcare costs. The incidence of VAP varies widely between 5% and 40%.[1] Neurocritical care patients are at an increased risk for VAP due to glottic dysfunction before intubation, the immunosuppressive effects secondary to brain injury, the need for frequent and multiple intrahospital shifts to the operating room and imaging room, the requirement of prolonged sedation for raised intracranial pressure, etc.[2,3] As VAP is an important cause of increased morbidity and mortality, accurate surveillance for VAP and benchmarking of hospitals is extremely important. THE PROBLEM WITH THE CLINICAL DEFINITION OF VENTILATOR-ASSOCIATED PNEUMONIA – OVERDIAGNOSIS AND UNDERDIAGNOSIS The traditional surveillance criteria for VAP were extremely subjective, relying heavily on clinical, radiographic, and microbiologic findings, often open to variability in the interpretation. This discrepancy potentially resulted in under-reporting or over-reporting of VAP, even among experts. A study by Tejerina et al. found that there was a significant discrepancy between clinically diagnosed VAP cases and their autopsy correlation.[4] Similarly, Nussenblatt et al. reported that experts rejected more than two-thirds of VAP cases identified by clinicians as they did not fulfill the diagnostic criteria.[5] This overdiagnosis leads to inappropriate use of broad-spectrum antibiotics, with pressure on antibiotic resistance. Another limitation of the Centers for Disease Control and Prevention’s (CDC’s) traditional VAP criteria is the tendency for hospitals to underreport cases when placed under scrutiny. Klompas noted a trend to interpret subjective criteria more strictly and demonstrate lower VAP rates, even when true clinical improvement had not occurred.[6] The most fundamental limitation of the CDC’s traditional VAP surveillance criteria is that they focus exclusively on VAP and thus fail to capture other complications, such as acute respiratory distress syndrome (ARDS) and atelectasis limiting their utility for improving overall patient outcomes. INTRODUCTION OF THE SURVEILLANCE BUNDLE – VENTILATOR-ASSOCIATED EVENTS In an attempt to address these limitations, in 2013, the CDC proposed new criteria that were supposed to be more objective and have a wider focus on other pulmonary complications such as ARDS, atelectasis, and pulmonary edema.[7] The development of the ventilator-associated events (VAEs) framework, which includes the concept of ventilator-associated conditions (VACs), was considered a significant move because it encouraged more objective criteria. While VAP remains a key focus of infection prevention efforts, the broader framework helps healthcare providers capture a more comprehensive picture of the challenges associated with mechanical ventilation. These definitions are simpler and amenable to automation, making the surveillance process more efficient, and reducing human error or bias. DOES THE NEW KID ON THE BLOCK OUTPERFORM THE OLD? Comparisons between the old and new criteria were always on, although the CDC categorically maintained that the new criteria are for surveillance only and not for clinical management. Muscedere et al. attempted to find the correlation between VAC, infection-related VAC (iVAC), and VAP and their clinical outcomes.[8] The data revealed that although VAC and iVAC picked up clinically diagnosed VAP cases, a significant number of patients with VAC and iVAC, as per the new definition, were not diagnosed with VAP. Conversely, only a minority of patients who were clinically diagnosed as VAP met the definition of VAC or iVAC. The authors attribute the discrepancy to the presence of a noninfectious cause of fever, alteration in white blood cell count to stress, a slight deterioration in parameters not reaching the thresholds for the diagnosis of VAC, absence of a period of clinical stability or improving ventilatory settings and possible inaccuracies in data collection [Figure 1]. Similar results suggesting a poor correlation between VAE and VAP have been suggested by others in the literature, with the sensitivity of VAE to detect clinical VAP being 42% and a positive predictive value of 23%.[9]Figure 1: The concordance between ventilator-associated events and clinically diagnosed ventilator-associated pneumonia. VAEs: Ventilator-associated Event, VAP: Ventilator-associated pneumoniaUNIQUE PROBLEMS IN DIAGNOSING VENTILATOR-ASSOCIATED EVENT AND VENTILATOR-ASSOCIATED PNEUMONIA IN NEUROCRITICAL CARE Application of the CDC’s surveillance criteria for VAEs, including VACs, may be suboptimal in neurocritical care settings for several reasons. These patients may develop VAP without exhibiting significant ventilator changes (e.g., increases in positive end-expiratory pressure [PEEP] or fraction of inspired oxygen [FiO2]) required by the CDC’s VAC definitions. The PEEP adjustments may often be discouraged by the treating team for fear of a rise in intracranial pressure and hemodynamic instability. Brain injury can blunt the fever response, and changes in ventilation may be minimal due to central nervous system depression. As a result, the objective criteria used to identify VAE events may not be triggered in neurocritical patients, leading to underdiagnosis of VAP in this population. The clinical diagnosis of VAP is also challenging as patients can have fever due to noninfectious causes, leukocytosis due to corticosteroids, stress response, etc., and are at a higher risk of aspiration before and during intubation.[10] The chest X-ray can also be deceptive, being an antero-posterior (AP) film, with numerous artifacts for abnormalities on chest X-ray. Similar findings were echoed in the study by Pallivalappil et al. in this issue of the Journal of Advanced Lung Health, where they published their findings from retrospective data over 3 months in a dedicated neurosurgery unit.[11] Using a standard clinical definition to define VAP, the investigators found a glaring incongruence between the definitions of VAE and VAP and an alarmingly increased prevalence of multidrug-resistant organisms.[11] While the authors should be congratulated meticulous analysis of the data presented, the study does not rule out the significance of routine use of the CDC surveillance criteria. Being a retrospective study, there can be errors in interpreting the available data as it might not depict the exact temporal factors. The clinical criteria of fever or hypothermia can have numerous noninfectious causes in a neurocritical care unit, which, on retrospective analysis, can be misclassified as infectious.[10] It needs to be clarified who supervised the ventilator settings of Peep and FiO2 what protocol was followed for setting them, and what baseline saturation was targeted in these cases. For instance, if the baseline saturation was maintained at 100% with a FiO2 or PEEP higher than what was required, the window for identifying a rise in FiO2 or PEEP can be lost. The fact that 64% of patients were already on “watch group” antibiotics before they had clinically proven VAP points out that a VAE was either missed out or there was inappropriate antibiotic use. The fact that microbiological isolation was available in the initial days (day 1–day 3) after ICU admission also indicates that these patients were at high risk of being colonized previously, as the earliest date of the event for VAE is day 3 of mechanical ventilation. This study points toward the need for dedicated physicians to care for critically ill patients, identifying infections early, distinguishing them from colonization, and avoiding unnecessary use of antibiotics. The authors also stress the need for a clinical diagnosis of VAP, which is extremely important for the timely administration of appropriate antibiotics. Timeliness is critical in managing VAP and waiting for clinical worsening invokes havoc. Patients with clinically suspected pneumonia but stable ventilator settings may be candidates for very short courses of antibiotics also.[12] VAE is not meant to be a real-time clinical diagnosis, as the sensitivity and specificity of VAP are low. The authors should be congratulated for the efficient use of the available data to stress the importance of clinical diagnosis versus surveillance definitions and for the efforts taken to find lacunae in the existing healthcare system, go back, take pains, and rectify them. As trained pulmonologists are well versed in reading X-rays and interpreting bedside signs, the clinical diagnosis of VAP and decision of usage of appropriate antibiotics should be ideally vested on them. VAE should be used only for surveillance, identifying the sickest patient with a high mortality risk. Such an approach will help in early and appropriate identification of sick patients, thus reducing the risk of antimicrobial resistance.