TY - GEN
T1 - Diagnosis and treatment of acutely hospitalised patients with suspected community -acquired pneumonia – clinical and microbiological perspectives
AU - Cartuliares, Mariana Bichuette
PY - 2023/10/19
Y1 - 2023/10/19
N2 - Community-acquired pneumonia (CAP) is a leading cause of mortalityworldwide, associated with high morbidity and hospital costs. CAP is one ofthe most common infections diagnosed in Emergency Departments (ED),requiring timely antibiotic treatment within a few hours from patientadmission. The CAP diagnosis is often based on uncertain history,questionable diagnostic methods and unspecific blood tests. This challengesclinicians to make correct early diagnoses, with misdiagnosis risking adverseevents, poorer patient outcomes, increased healthcare costs and the overuseof broad-spectrum antibiotics. This, in turn, contributes to the increaseddevelopment of resistant bacteria, thereby threatening future treatmentpossibilities.This PhD thesis aimed to investigate potential improvements to the CAP diagnostic process within the first hours of acute admission. An improved diagnostic process would use rapid and precise diagnosis methods to support rational, targeted antibiotic prescriptions, preventing poor patient outcomes and antimicrobial resistance (AMR).In study I, it was hypothesised that well-defined clinical characteristics could assist ED physicians in making an earlier, more accurate CAP diagnosis. The study design was a cross-sectional diagnostic, predictive study. The study identified the clinical characteristics of patients with CAP, developed a diagnostic model and compared the model’s performance to the ED physician’s initial assessment. The model yielded 13 predictors, all recognised and supported by published literature. The performance and calibration of the model were good but did not outperform the initial tentative diagnosis made by the ED physicians. The addition of new diagnostic tools will be essential in future diagnostic models.In study II, it was hypothesised that expiratory techniques (forced expiratory technique and sputum induction) were non-inferior to tracheal suction for collecting good-quality sputum samples from patients with suspected lower respiratory tract (LRT) infection in the ED. The number of adverse events between groups was compared, and patient experiences of sampling methods were investigated. For this thesis, additonal microbiological results of goodquality LRT specimens are described. The study was an open-label, parallelarmed, non-inferiority randomised controlled trial (RCT). Results showed that tracheal suction had approximately a twofold likelihood of ensuring a good-quality specimen compared with expiratory techniques. However, often good-quality samples had low microbiological yields. Although there were no differences when adverse events between the two sampling methods were pooled and compared, patients allocated to the expiratory technique reported a more positive experience than patients allocated to tracheal suction.In study III, it was hypothesised that point-of-care polymerase chain reaction (POC-PCR) testing of LRT samples from suspected CAP patients would increase the proportion of patients treated with no or narrow-spectrum antibiotics compared with standard care only (SCO), which included routine culture and targeted-specific PCR if requested by the ED physician. The study compared the length of stay (LOS), intensive care unit (ICU) admission, mortality and readmissions between groups. Additional descriptive analysis was completed on bacteria and viruses from the microbiological analyses of the LRT specimens. The study was a multicentre, open-label, parallel-armed superiority RCT. Adding POC-PCR to the diagnostic setup did not increase the number of patients treated with narrow-spectrum or without antibiotics, but the results indicated that patients in the POC-PCR group received earlier and more targeted antibiotic treatments. Compared with culture, POC-PCR identified more bacteria and viruses, including common CAP pathogens. Nostatistical differences between POC-PCR and SCO groups were observed for mortality, readmissions, ICU admissions or LOS.In conclusion, this thesis reflects the challenges of diagnosing CAP and provides new insight into optimising the diagnostic process. These three studies contribute vital information and knowledge to future research and implementation strategies targeting the improvement of CAP diagnosis.
AB - Community-acquired pneumonia (CAP) is a leading cause of mortalityworldwide, associated with high morbidity and hospital costs. CAP is one ofthe most common infections diagnosed in Emergency Departments (ED),requiring timely antibiotic treatment within a few hours from patientadmission. The CAP diagnosis is often based on uncertain history,questionable diagnostic methods and unspecific blood tests. This challengesclinicians to make correct early diagnoses, with misdiagnosis risking adverseevents, poorer patient outcomes, increased healthcare costs and the overuseof broad-spectrum antibiotics. This, in turn, contributes to the increaseddevelopment of resistant bacteria, thereby threatening future treatmentpossibilities.This PhD thesis aimed to investigate potential improvements to the CAP diagnostic process within the first hours of acute admission. An improved diagnostic process would use rapid and precise diagnosis methods to support rational, targeted antibiotic prescriptions, preventing poor patient outcomes and antimicrobial resistance (AMR).In study I, it was hypothesised that well-defined clinical characteristics could assist ED physicians in making an earlier, more accurate CAP diagnosis. The study design was a cross-sectional diagnostic, predictive study. The study identified the clinical characteristics of patients with CAP, developed a diagnostic model and compared the model’s performance to the ED physician’s initial assessment. The model yielded 13 predictors, all recognised and supported by published literature. The performance and calibration of the model were good but did not outperform the initial tentative diagnosis made by the ED physicians. The addition of new diagnostic tools will be essential in future diagnostic models.In study II, it was hypothesised that expiratory techniques (forced expiratory technique and sputum induction) were non-inferior to tracheal suction for collecting good-quality sputum samples from patients with suspected lower respiratory tract (LRT) infection in the ED. The number of adverse events between groups was compared, and patient experiences of sampling methods were investigated. For this thesis, additonal microbiological results of goodquality LRT specimens are described. The study was an open-label, parallelarmed, non-inferiority randomised controlled trial (RCT). Results showed that tracheal suction had approximately a twofold likelihood of ensuring a good-quality specimen compared with expiratory techniques. However, often good-quality samples had low microbiological yields. Although there were no differences when adverse events between the two sampling methods were pooled and compared, patients allocated to the expiratory technique reported a more positive experience than patients allocated to tracheal suction.In study III, it was hypothesised that point-of-care polymerase chain reaction (POC-PCR) testing of LRT samples from suspected CAP patients would increase the proportion of patients treated with no or narrow-spectrum antibiotics compared with standard care only (SCO), which included routine culture and targeted-specific PCR if requested by the ED physician. The study compared the length of stay (LOS), intensive care unit (ICU) admission, mortality and readmissions between groups. Additional descriptive analysis was completed on bacteria and viruses from the microbiological analyses of the LRT specimens. The study was a multicentre, open-label, parallel-armed superiority RCT. Adding POC-PCR to the diagnostic setup did not increase the number of patients treated with narrow-spectrum or without antibiotics, but the results indicated that patients in the POC-PCR group received earlier and more targeted antibiotic treatments. Compared with culture, POC-PCR identified more bacteria and viruses, including common CAP pathogens. Nostatistical differences between POC-PCR and SCO groups were observed for mortality, readmissions, ICU admissions or LOS.In conclusion, this thesis reflects the challenges of diagnosing CAP and provides new insight into optimising the diagnostic process. These three studies contribute vital information and knowledge to future research and implementation strategies targeting the improvement of CAP diagnosis.
U2 - 10.21996/2tx7-q194
DO - 10.21996/2tx7-q194
M3 - Ph.D. thesis
PB - Syddansk Universitet. Det Sundhedsvidenskabelige Fakultet
ER -