Diagnosing pneumonia in primary care : Aspects of the value of clinical and laboratory findings and the use of chest X-ray

Sammanfattning: It is important to identify patients with pneumonia because it is potentially a serious disease, often of bacterial origin, that should be treated with antibiotics. It is equally important to identify those with acute bronchitis, a self-limiting disease, that should not be treated with antibiotics. Because bacterial resistance is increasing, over-prescribing of antibiotics should be avoided. However, it is sometimes difficult to differentiate between the two diagnoses, and guidelines concerning the assessment do not conform.  The general aim of this thesis was to investigate if diagnostics of pneumonia in primary care can be improved and whether this could contribute to reduced prescription of antibiotics.  As a first step, different anamnestic, clinical and laboratory findings and the doctor’s degree of suspicion of pneumonia in primary care were compared with chest X-ray (CXR) findings. The doctor’s degree of suspicion of pneumonia was shown to be a good predictor. When the physician was sure of the diagnosis, the likelihood for radiographic pneumonia was high and when quite sure, CXR was positive in less than half of the cases.  To further improve the diagnostics of pneumonia, and thus reduce antibiotic prescriptions, patients were referred for CXR when the physician was unsure or quite sure of a pneumonia diagnosis. The intervention did not result in any decrease in antibiotic prescriptions compared with a control group. However, it emerged that the physicians did not fully trust the CXR outcome, but prescribed antibiotics even when the results were negative.  To gain insight into the contribution of C-reactive protein (CRP) levels to the degree of suspicion, physicians were asked to estimate their degree of suspicion of pneumonia before and after CRP testing. CRP affected the degree of suspicion to a great extent, and most often resulted in a lowered degree of suspicion and thereby in the clinical decision of dismissing the diagnosis of pneumonia.  The use of different diagnostic tests and prescription of antibiotics in the assessment of acute bronchitis and pneumonia over time was evaluated in a register-based study. The study showed that the use of diagnostic tests for both diagnoses has increased, and that there has been a reduction in antibiotic prescriptions for acute bronchitis.  In conclusion, the doctor’s degree of suspicion of pneumonia seems to be a good predictor of the condition. When the physician is sure of the diagnosis, no further investigation is needed, and antibiotics can be prescribed on reliable grounds. CRP testing affects the degree of suspicion and is most valuable when unsure of the diagnosis where it can be helpful to exclude pneumonia. In contrast, more extensive use of CXR does not contribute to a decrease in antibiotic prescriptions in the diagnostics of pneumonia.