Impact on the Use of Antibiotics of a Multimodal Algorithm for the Diagnosis and Management of Acute Community-acquired Pneumonia in the Emergency Room

Acute community-acquired pneumonia (CAP) is the leading infectious cause of infectious death worldwide and the third leading cause of death from all causes. They are also very frequent reasons for visit to the emergency room, especially during winter season. Although respiratory viruses are responsible for approximately 30 to 50% of CAP, antibiotics' prescription remain very high. This prescription is at the origin of the phenomenon of antibiotic resistance, which continues to increase throughout the world. In the meantime, the biological testing of this clinical picture is increasingly evolving since a decade, and this evolution has been even accelerated with the emergence of COVID-19. With regard to the increasing availability of respiratory viruses testing, we need more studies allowing to better use these results to spare antibiotics when useless.

Several avenues of study have been investigated to improve the diagnosis of bacterial CAP and thus reduce unnecessary antibiotic therapy: differentiating viral CAP from bacterial CAP using multiplex PCR and/or inflammation biomarkers, localizing the infection lung parenchyma using chest CT.

The various studies carried out on the impact of the use of multiplex PCR in the emergency department (ED) led to discordant conclusions. The study carried out within Bichat Claude-Bernard Hospital only shows an interest for single room assignment in patients infected with pathogens such as influenza, RSV and Metapneumovirus. Multiplex PCR delocalized to the ED does not seem to be a sufficient measure to reduce the prescription of antibiotics in patients suspected of CAP admitted from the emergency room. One study carried during the COVID-19 pandemic highlighted that multiplex PCR, used as a point of care, induced improvements in patient flow and infection control measures.

C Reactive Protein (CRP) is a well-known inflammation biomarker. Previous studies estimate that CRP has better performance than procalcitonin in the diagnosis of pneumonia with thresholds often described at 50 mg.L-1 and revised upwards in older patients. The National Institute for Health and Care Excellence (NICE) has endorsed the use of point-of-care CRP to diagnose CAP and reduce inappropriate antibiotic use. This recent meta-analysis suggests that CRP is a better marker than PCT for the diagnosis of pneumonia. Indeed, a CRP >20 mg/L, >50 mg/L or >100 mg/L has a positive LR+ likelihood ratio of 2.08, 3.68, and 5.79, respectively, and a negative LR- likelihood ratio of 0.32, 0.36, and 0.48, while PCT >0.25 µg/L or >0.50 µg/L has an LR+ of 5.43 and 8.25, respectively, and an LR- of 0.62 and 0.76.

In addition, the place of imaging in reducing the prescription of antibiotics in CAP is still very little studied; nevertheless, CT without injection has proven to be of great help in the diagnosis of COVID-19 and remains the gold standard (reference examination) for effectively diagnosing CAP.

The current recommendations of the French society of infectious diseases do not clearly mention the use of multiplex PCR. CRP is included as a biomarker of inflammation in favor of bacterial infection when it is high. The CT scan is mentioned in cases where the diagnosis of CAP is difficult without defining the criteria for this difficult diagnosis. The College of Pneumology Teachers updated in 2020 places multiplex PCR as an essential diagnostic tool in the management of CAP. Finally, the latest American recommendations (2018) mention the performance of a multiplex PCR as indicated in any patient admitted to hospital with symptoms of influenza-like illness. As a result, the practices of hospitals are very heterogeneous in the diagnosis of CAP, we therefore propose to develop a multimodal algorithm combining multiplex PCR, CRP and chest CT scan in the diagnosis of CAP in patients requiring hospitalization, starting as soon as possible during winter.